Automated focusing in bright-field microscopy for tuberculosis detection

Phase flicker-induced sharpness deterioration on 2D holographic displays with digitally driven phase-only LCoS devices

Phase flicker in liquid crystal on silicon (LCoS) devices can decrease the effective phase modulation resolution by introducing overlapped phase oscillations between adjacent modulated gray levels, thus degrading the performance of LCoS devices in various applications. However, the effect of phase flicker on a holographic display is often overlooked. From an application angle, this paper investigates the quality of the holographic reconstructed image, especially sharpness, under the static and dynamic effects of different flicker magnitudes. Both the simulation and experimental results reveal that the increment in the magnitude of phase flicker causes an equal sharpness deterioration with the reduction of the numbers of hologram phase modulation levels.

Computational techniques for the automated detection of mycobacterium tuberculosis from digitized sputum smear microscopic images: A systematic review

BACKGROUND

Tuberculosis is an infectious disease that is caused by Mycobacterium tuberculosis (MTB), which mostly affects the lungs of humans. Bright-field microscopy and fluorescence microscopy are two major testing techniques used for tuberculosis (TB) detection. TB bacilli were identified and counted manually from sputum under a microscope and were found to be tedious, laborious and error prone. To eliminate this problem, traditional image processing techniques and deep learning (DL) models were deployed here to build computer-aided diagnosis (CADx) systems for TB detection.

METHODS

In this paper, we performed a systematic review on image processing techniques used in developing computer-aided diagnosis systems for TB detection. Articles selected for this review were retrieved from publication databases such as Science Direct, ACM, IEEE Xplore, Springer Link and PubMed. After a rigorous pruning exercise, 42 articles were selected, of which 21 were journal articles and 21 were conference articles.

RESULT

Image processing techniques and deep neural networks such as CNN and DCNN proposed in the literature along with clinical applications are presented and discussed. The performance of these techniques has been evaluated on metrics such as accuracy, sensitivity, specificity, precision and F-1 score and is presented accordingly.

CONCLUSION

CADx systems built on DL models performed better in TB detection and classification due to their abstraction of low-level features, better generalization and minimal or no human intervention in their operations. Research gaps identified in the literature have been highlighted and discussed for further investigation.

An algorithm selection methodology for automated focusing in optical microscopy

Autofocus systems are essential in optical microscopy. These systems typically sweep the sample through the focal range and apply an algorithm to determine the contrast value of each image, where the highest value indicates the optimal focus position. As the optimal algorithm may vary according to the images' content, we evaluate the 15 most used algorithms in the field using 150 stacks of images from four different kinds of tissue. We use four measuring criteria and two types of analysis and propose a general methodology to apply to select the best fitting algorithm for any given application. In this paper, we present the results of this evaluation and a detailed discussion of different features: the threshold used for the algorithms, the criteria parameters, the analysis used, the bit depth of the images, their magnification, and the type of tissue, reaching the conclusion that some of these parameters are more relevant to the study than others, and the implementation of the proposed methodology can lead to a fast and reliable autofocus system capable of performing an analysis and selection of algorithms with no supervision required.

Recent technical advances in whole slide imaging instrumentation

Microscopic observation of biological specimen smears is the mainstay of diagnostic pathology, as defined by the Digital Pathology Association. Though automated systems for this are commercially available, their bulky size and high cost renders them unusable for remote areas. The research community is investing much effort towards building equivalent but portable, low-cost systems. An overview of such research is presented here, including a comparative analysis of recent reports. This paper also reviews recently reported systems for automated staining and smear formation, including microfluidic devices; and optical and computational automated microscopy systems including smartphone-based devices. Image pre-processing and analysis methods for automated diagnosis are also briefly discussed. It concludes with a set of foreseeable research directions that could lead to affordable, integrated and accurate whole slide imaging systems.

Neoteric advancements in TB diagnostics and its future frame

Tuberculosis (TB) is one of the major infectious disease that causes threat to human health and leads to death in most of the cases. Mycobacterium tuberculosis is the causative agent that can affect both pulmonary and extra pulmonary regions of the body. This infection can be presented either as an active or latent form in the patients. Although this disease has been declared curable and preventable by WHO, it still holds its position as a global emergency. Over the past decade many hurdles such as low immunity, co-infections like HIV, autoimmune disorders, poverty, malnutrition and emerging trends in drug resistance patterns are hindering the eradication of this infection. However, many programmes have been launched by WHO with involvement of governments at various level to put a full stop over the disease. Under the Revised National Tuberculosis Control Programme (RNTCP) which was recently renamed as National Tuberculosis Elimination Programme (NTEP), the major focus is on eliminating tuberculosis by the year 2025. The main aim of the programme is to identify feasible quality testing, evaluate through NIKSHYA poshak yozana, restrict through BCG vaccination and assemble with public awareness to eradicate MTB. Numerous novel diagnostic techniques and molecular tools have been developed to elucidate and differentiate report of various suspected and active tuberculosis patients. However, improvements are still required to cut short the duration of the overall process ranging from screening of patients to their successful treatment.

Artificial Intelligence-Based Screening for Mycobacteria in Whole-Slide Images of Tissue Samples

OBJECTIVES

This study aimed to develop and validate a deep learning algorithm to screen digitized acid fast-stained (AFS) slides for mycobacteria within tissue sections.

METHODS

A total of 441 whole-slide images (WSIs) of AFS tissue material were used to develop a deep learning algorithm. Regions of interest with possible acid-fast bacilli (AFBs) were displayed in a web-based gallery format alongside corresponding WSIs for pathologist review. Artificial intelligence (AI)-assisted analysis of another 138 AFS slides was compared to manual light microscopy and WSI evaluation without AI support.

RESULTS

Algorithm performance showed an area under the curve of 0.960 at the image patch level. More AI-assisted reviews identified AFBs than manual microscopy or WSI examination (P < .001). Sensitivity, negative predictive value, and accuracy were highest for AI-assisted reviews. AI-assisted reviews also had the highest rate of matching the original sign-out diagnosis, were less time-consuming, and were much easier for pathologists to perform (P < .001).

CONCLUSIONS

This study reports the successful development and clinical validation of an AI-based digital pathology system to screen for AFBs in anatomic pathology material. AI assistance proved to be more sensitive and accurate, took pathologists less time to screen cases, and was easier to use than either manual microscopy or viewing WSIs.

Transmissive Single-Pixel Microscopic Imaging through Scattering Media

Microscopic imaging is of great significance for medical diagnosis. However, due to the strong scattering and absorption of tissue, the implementation of non-invasive microscopic imaging is very difficult. Traditional single-pixel microscopes, based on reflective optical systems, provide an alternative solution for scattering media imaging. Here, the single-pixel microscope with transmissive liquid crystal modulation is proposed. The microscopic ability of the proposed microscope is calibrated. The multi-spectral microscopic imaging of the object is demonstrated. The transmissive imaging of the object behind the scattering media is analyzed. The proposed prototype of the transmissive single-pixel microscope is expected to be applied in microscopic imaging through scattering media and medical imaging.

FrenchFISH: Poisson Models for Quantifying DNA Copy Number From Fluorescence In Situ Hybridization of Tissue Sections

PURPOSE

Chromosomal aberration and DNA copy number change are robust hallmarks of cancer. The gold standard for detecting copy number changes in tumor cells is fluorescence in situ hybridization (FISH) using locus-specific probes that are imaged as fluorescent spots. However, spot counting often does not perform well on solid tumor tissue sections due to partially represented or overlapping nuclei.

MATERIALS AND METHODS

To overcome these challenges, we have developed a computational approach called FrenchFISH, which comprises a nuclear volume correction method coupled with two types of Poisson models: either a Poisson model for improved manual spot counting without the need for control probes or a homogeneous Poisson point process model for automated spot counting.

RESULTS

We benchmarked the performance of FrenchFISH against previous approaches using a controlled simulation scenario and tested it experimentally in 12 ovarian carcinoma FFPE-tissue sections for copy number alterations at three loci (c-Myc, hTERC, and SE7). FrenchFISH outperformed standard spot counting with 74% of the automated counts having < 1 copy number difference from the manual counts and 17% having < 2 copy number differences, while taking less than one third of the time of manual counting.

CONCLUSION

FrenchFISH is a general approach that can be used to enhance clinical diagnosis on sections of any tissue by both speeding up and improving the accuracy of spot count estimates.

Autofocusing technologies for whole slide imaging and automated microscopy

Whole slide imaging (WSI) has moved digital pathology closer to diagnostic practice in recent years. Due to the inherent tissue topography variability, accurate autofocusing remains a critical challenge for WSI and automated microscopy systems. The traditional focus map surveying method is limited in its ability to acquire a high degree of focus points while still maintaining high throughput. Real-time approaches decouple image acquisition from focusing, thus allowing for rapid scanning while maintaining continuous accurate focus. This work reviews the traditional focus map approach and discusses the choice of focus measure for focal plane determination. It also discusses various real-time autofocusing approaches including reflective-based triangulation, confocal pinhole detection, low-coherence interferometry, tilted sensor approach, independent dual sensor scanning, beam splitter array, phase detection, dual-LED illumination and deep-learning approaches. The technical concepts, merits and limitations of these methods are explained and compared to those of a traditional WSI system. This review may provide new insights for the development of high-throughput automated microscopy imaging systems that can be made broadly available and utilizable without loss of capacity.

An adaptable analysis workflow for characterization of platelet spreading and morphology

The assessment of platelet spreading through light microscopy, and the subsequent quantification of parameters such as surface area and circularity, is a key assay for many platelet biologists. Here we present an analysis workflow which robustly segments individual platelets to facilitate the analysis of large numbers of cells while minimizing user bias. Image segmentation is performed by interactive learning and touching platelets are separated with an efficient semi-automated protocol. We also use machine learning methods to robustly automate the classification of platelets into different subtypes. These adaptable and reproducible workflows are made freely available and are implemented using the open-source software KNIME and ilastik.

Parallel implementations to accelerate the autofocus process in microscopy applications

Several autofocus algorithms based on the analysis of image sharpness have been proposed for microscopy applications. Since autofocus functions (AFs) are computed from several images captured at different lens positions, these algorithms are considered computationally intensive. With the aim of presenting the capabilities of dedicated hardware to speed-up the autofocus process, we discuss the implementation of four AFs using, respectively, a multicore central processing unit (CPU) architecture and a graphic processing unit (GPU) card. Throughout different experiments performed on 300 image stacks previously identified with tuberculosis bacilli, the proposed implementations have allowed for the acceleration of the computation time for some AFs up to 23 times with respect to the serial version. These results show that the optimal use of multicore CPU and GPUs can be used effectively for autofocus in real-time microscopy applications.

Adaptive predictive scanning method based on a high-precision automatic microscopy system

Predicting the focal plane is an effective method to increase the scanning speed of an automatic microscopy system. However, the image easily defocuses when using traditional predictive scanning methods. In this paper, we introduce an adaptive predictive scanning method (APSM) that greatly improves the accuracy of predictive scanning. Instead of using a fixed planar model to predict the focal plane position, APSM updates the predicted focal plane in real time based on the focal position of the reference point during the scanning process, thus predicting the focal position of each local view more accurately. Using the APSM, the average image defocus value is 0.39 μm, while conventional predictive scanning methods reach 1.05 μm. APSM greatly improves focal accuracy and can be applied to a high-precision automatic microscopy system.

Ring Difference Filter for Fast and Noise Robust Depth from Focus

Depth from focus (DfF) is a method of estimating the depth of a scene by using information acquired through changes in the focus of a camera. Within the DfF framework of, the focus measure (FM) forms the foundation which determines the accuracy of the output. With the results from the FM, the role of a DfF pipeline is to determine and recalculate unreliable measurements while enhancing those that are reliable. In this paper, we propose a new FM, which we call the "ring difference filter" (RDF), that can more accurately and robustly measure focus. FMs can usually be categorized as confident local methods or noise robust non-local methods. The RDF's unique ring-and-disk structure allows it to have the advantages of both local and non-local FMs. We then describe an efficient pipeline that utilizes the RDF's properties. Part of this pipeline is our proposed RDF-based cost aggregation method, which is able to robustly refine the initial results in the presence of image noise. Our method is able to reproduce results that are on par with or even better than those of state-of-the-art methods, while spending less time in computation.

Automatic microscopic detection of mycobacteria in sputum: a proof-of-concept

The laboratory diagnosis of lung mycobacterioses including tuberculosis comprises the microscopic examination of sputum smear after appropriate staining such as Ziehl-Neelsen staining to observe acid-fast bacilli. This standard procedure is operator-dependant and its sensitivity depends on the duration of observation. We developed and evaluated an operator-independent microscopic examination of sputum smears for the automated detection and enumeration of acid-fast bacilli using a ZEISS Axio Scan.Z1 microscope. The sensitivity, specificity, positive predictive value, negative predictive values and accuracy were calculated using standard formulations by comparison with standard microscopic examination. After in-house parameterization of the automatic microscope and counting software, the limit of detection evaluated by seeding negative sputa with Mycobacterium bovis BCG or Mycobacterium tuberculosis H37Rv (10⁰–10⁵ bacilli/mL) was of 10² bacilli/mL of sputum with a 100% positivity rate. Then, the evaluation of 93 sputum specimens including 34 smear-positive and 59 smear-negative specimens yielded a sensitivity of 97.06% [84.67–99.93%], a specificity of 86.44% [73.01–92.78%]. Up to 100 smear slides could be stocked for reading in the microscope magazine and results are exportable into the laboratory information system. Based on these preliminary results, we are implanting this automatic protocol in the routine workflow so that only smears detected positive by automatic microscopy are confirmed by standard microscopic examination.

Continuous stacking computational approach based automated microscope slide scanner

Cost-effective and automated acquisition of whole slide images is a bottleneck for wide-scale deployment of digital pathology. In this article, a computation augmented approach for the development of an automated microscope slide scanner is presented. The realization of a prototype device built using inexpensive off-the-shelf optical components and motors is detailed. The applicability of the developed prototype to clinical diagnostic testing is demonstrated by generating good quality digital images of malaria-infected blood smears. Further, the acquired slide images have been processed to identify and count the number of malaria-infected red blood cells and thereby perform quantitative parasitemia level estimation. The presented prototype would enable cost-effective deployment of slide-based cyto-diagnostic testing in endemic areas.

Establishment of hybridized focus measure functions as a universal method for autofocusing

Exact focusing is essential for any automatic image capturing system. Performances of focus measure functions (FMFs) used for autofocusing are sensitive to image contents and imaging systems. Therefore, identification of universal FMF assumes a lot of significance. Eight FMFs were hybridized in pairs of two and implemented simultaneously on a single stack to calculate the hybrid focus measure. In total, 28 hybrid FMFs (HFMFs) and eight FMFs were implemented on stacks of images from three different imaging modalities. Performance of FMFs was found to the best at 50% region sampling. Accuracy, focus error, and false maxima were calculated to evaluate the performance of each FMF. Nineteen HFMFs provided >90% accuracy. Image distortion (noise, contrast, saturation, illumination, etc.) was performed to evaluate robustness of HFMFs. Hybrid of tenengrad variance and steerable filter-based (VGRnSFB) FMFs was identified as the most robust and accurate function with an accuracy of ≥90% and a relatively lower focus error and false maxima rate. Sharpness of focus curve of VGRnSFB along with eight individual FMFs was also computed to determine the efficacy of HFMF for the optimization process. VGRnSFB HFMF may be implemented for automated capturing of an image for any imaging system.

Ziehl-Neelsen sputum smear microscopy image database: a resource to facilitate automated bacilli detection for tuberculosis diagnosis

Ziehl-Neelsen stained microscopy is a crucial bacteriological test for tuberculosis detection, but its sensitivity is poor. According to the World Health Organization (WHO) recommendation, 300 viewfields should be analyzed to augment sensitivity, but only a few viewfields are examined due to patient load. Therefore, tuberculosis diagnosis through automated capture of the focused image (autofocusing), stitching of viewfields to form mosaics (autostitching), and automatic bacilli segmentation (grading) can significantly improve the sensitivity. However, the lack of unified datasets impedes the development of robust algorithms in these three domains. Therefore, the Ziehl-Neelsen sputum smear microscopy image database (ZNSM iDB) has been developed, and is freely available. This database contains seven categories of diverse datasets acquired from three different bright-field microscopes. Datasets related to autofocusing, autostitching, and manually segmenting bacilli can be used for developing algorithms, whereas the other four datasets are provided to streamline the sensitivity and specificity. All three categories of datasets were validated using different automated algorithms. As images available in this database have distinctive presentations with high noise and artifacts, this referral resource can also be used for the validation of robust detection algorithms. The ZNSM-iDB also assists for the development of methods in automated microscopy.

Autofocusing system for spatial light modulator-based maskless lithography

To produce diffractive or holographic structures in a photolithographic process, an optical projection system enabling structure resolution in the submicrometer range is highly desirable. To ensure that the optical focus of such a system lies on the substrate surface during the whole lithographic fabrication process, an autofocus system able to focus on a depth of field of a few hundred nanometers is usually required. In this work, we developed an autofocus system for spatial light modulator (SLM)-based maskless photolithographic applications. The system is capable of high-precision focusing without affecting the photoresist performance. It is based on contrast measurement combined with focus-pattern illumination to ensure high contrast at the substrate surface. In addition, we evaluated various autofocus algorithms with respect to time efficiency and accuracy to determine suitable focus-pattern and focus-algorithm combinations.

FocusALL: Focal Stacking of Microscopic Images Using Modified Harris Corner Response Measure

Automated image analysis of microscopic images such as protein crystallization images and cellular images is one of the important research areas. If objects in a scene appear at different depths with respect to the camera's focal point, objects outside the depth of field usually appear blurred. Therefore, scientists capture a collection of images with different depths of field. Focal stacking is a technique of creating a single focused image from a stack of images collected with different depths of field. In this paper, we introduce a novel focal stacking technique, FocusALL, which is based on our modified Harris Corner Response Measure. We also propose enhanced FocusALL for application on images collected under high resolution and varying illumination. FocusALL resolves problems related to the assumption that focus regions have high contrast and high intensity. Especially, FocusALL generates sharper boundaries around protein crystal regions and good in focus images for high resolution images in reasonable time. FocusALL outperforms other methods on protein crystallization images and performs comparably well on other datasets such as retinal epithelial images and simulated datasets.

Compact multi-band fluorescent microscope with an electrically tunable lens for autofocusing

Autofocusing is a routine technique in redressing focus drift that occurs in time-lapse microscopic image acquisition. To date, most automatic microscopes are designed on the distance detection scheme to fulfill the autofocusing operation, which may suffer from the low contrast of the reflected signal due to the refractive index mismatch at the water/glass interface. To achieve high autofocusing speed with minimal motion artifacts, we developed a compact multi-band fluorescent microscope with an electrically tunable lens (ETL) device for autofocusing. A modified searching algorithm based on equidistant scanning and curve fitting is proposed, which no longer requires a single-peak focus curve and then efficiently restrains the impact of external disturbance. This technique enables us to achieve an autofocusing time of down to 170 ms and the reproductivity of over 97%. The imaging head of the microscope has dimensions of 12 cm × 12 cm × 6 cm. This portable instrument can easily fit inside standard incubators for real-time imaging of living specimens.

Optimal self-calibration of tomographic reconstruction parameters in whole-body small animal optoacoustic imaging

In tomographic optoacoustic imaging, multiple parameters related to both light and ultrasound propagation characteristics of the medium need to be adequately selected in order to accurately recover maps of local optical absorbance. Speed of sound in the imaged object and surrounding medium is a key parameter conventionally assumed to be uniform. Mismatch between the actual and predicted speed of sound values may lead to image distortions but can be mitigated by manual or automatic optimization based on metrics of image sharpness. Although some simple approaches based on metrics of image sharpness may readily mitigate distortions in the presence of highly contrasting and sharp image features, they may not provide an adequate performance for smooth signal variations as commonly present in realistic whole-body optoacoustic images from small animals. Thus, three new hybrid methods are suggested in this work, which are shown to outperform well-established autofocusing algorithms in mouse experiments in vivo.

Autofocus using adaptive prediction approximation combined search for the fluorescence microscope in second-generation DNA sequencing system

Autofocus is an important technique for high-speed image acquisition in the second-generation DNA sequencing system, and this paper studies the passive focus algorithm for the system, which consists of two parts: focus measurement (FM) and focus search (FS). Based on the properties of DNA chips' images, we choose the normalized variance as the FM algorithm and develop a new robust FS named adaptive prediction approximation combined search (APACS). APACS utilizes golden section search (GSS) to approximate the focus position and engages the curve-fitting search (CFS) to predict the position simultaneously in every step of GSS. When the difference between consecutive predictions meets the set precision, the search finishes. Otherwise, it ends as GSS. In APACS, we also propose an estimation method, named the combination of centroid estimation and overdetermined equations estimation by least squares solution, to calculate the initial vector for the nonlinear equations in APACS prediction, which reduces the iterations and accelerates the search. The simulation and measured results demonstrate that APACS not only maintains the stability but also reduces the focus time compared with GSS and CFS, which indicates APACS is a robust and fast FS for the fluorescence microscope in a sequencing system.

Influence of the phase effect on gradient-based and statistics-based focus measures in bright field microscopy

Autofocusing is essential to high throughput microscopy and live cell imaging and requires reliable focus measures. Phase objects such as separated single Chinese hamster ovary cells are almost invisible at the optical focus position in bright field microscopy images. Because of the phase effect, defocused images of phase objects have more contrast. In this paper, we show that widely used focus measures exhibit an untypical behaviour for such images. In the case of homogeneous cells, that is, when most cells tend to lie in the same focal plane, both gradient-based and statistics-based focus measures tend to have a local minimum instead of a global maximum at the optical focus position. On the other hand, if images show inhomogeneous cells, gradient-based focus measures tend to yield typical focus curves, whereas statistics-based focus measures deliver curves similar to the case of homogeneous cells. These results were interpreted using the equation describing the phase effect and patch-wise analysis of the focus curves. Bioprocess engineering experts are also influenced by the phase effect. Forty-four focus positions selected by them led to the conclusion that they prefer to look at defocused images instead of those at the optical focus.

Autofocusing for Microscopic Images using Harris Corner Response Measure

One of the difficulties for proper imaging in microscopic image analysis is defocusing. Microscopic images such as cellular images, protein images, etc. need properly focused image for image analysis. A small difference in focal depth affects the details of an object significantly. In this paper, we introduce a novel auto-focusing approach based on Harris Corner Response Measure (HCRM) and compare the performance with some existing auto-focusing methods. We perform our experiments on protein images as well as a simulated image stack to evaluate the performance of our method. Our results show that our HCRM-based technique outperforms other techniques.

A sputum smear microscopy image database for automatic bacilli detection in conventional microscopy

In this work, we present an image database for automatic bacilli detection in sputum smear microscopy. The database comprises two parts. The first one, called the autofocus database, contains 1200 images with resolution of 2816 × 2112 pixels. This database was obtained from 12 slides, with 10 fields per slide. Each stack is composed of 10 images, with the fifth image in focus. The second one, called the segmentation and classification database, contains 120 images with resolution of 2816×2112 pixels. This database was obtained from 12 slices, with 10 fields per slice. In both databases, the images were acquired from fields of slides stained with the standard Kinyoun method. In both databases, accordingly to the background content, the images were classified as belonging to high background content or low background content. In all 120 images of segmentation and classification database, the identified objects were enclosed within a geometric shape by a trained technician. A true bacillus was enclosed in a circle. An agglomerated bacillus was enclosed by a rectangle and a doubtful bacillus (the image focus or geometry does not allow a clear identification of the object) was enclosed by a polygon. These marked objects could be used as a gold standard to calculate the accuracy, sensitivity and specificity of bacilli recognition.

Automatic section thickness determination using an absolute gradient focus function

Quantitative analysis of microstructures using computerized stereology systems is an essential tool in many disciplines of bioscience research. Section thickness determination in current nonautomated approaches requires manual location of upper and lower surfaces of tissue sections. In contrast to conventional autofocus functions that locate the optimally focused optical plane using the global maximum on a focus curve, this study identified by two sharp 'knees' on the focus curve as the transition from unfocused to focused optical planes. Analysis of 14 grey-scale focus functions showed, the thresholded absolute gradient function, was best for finding detectable bends that closely correspond to the bounding optical planes at the upper and lower tissue surfaces. Modifications to this function generated four novel functions that outperformed the original. The 'modified absolute gradient count' function outperformed all others with an average error of 0.56 μm on a test set of images similar to the training set; and, an average error of 0.39 μm on a test set comprised of images captured from a different case, that is, different staining methods on a different brain region from a different subject rat. We describe a novel algorithm that allows for automatic section thickness determination based on just out-of-focus planes, a prerequisite for fully automatic computerized stereology.

Creating a virtual slide map from sputum smear images for region-of-interest localisation in automated microscopy

We address the location of regions-of-interest in previously scanned sputum smear slides requiring re-examination in automated microscopy for tuberculosis (TB) detection. We focus on the core component of microscope auto-positioning, which is to find a point of reference, position and orientation, on the slide so that it can be used to automatically bring desired fields to the field-of-view of the microscope. We use virtual slide maps together with geometric hashing to localise a query image, which then acts as the point of reference. The true positive rate achieved by the algorithm was above 88% even for noisy query images captured at slide orientations up to 26°. The image registration error, computed as the average mean square error, was less than 14 pixel² (corresponding to 1.02 μm²). The algorithm is inherently robust to changes in slide orientation and placement and showed high tolerance to illumination changes and robustness to noise.

Comparative evaluation of autofocus algorithms for a real-time system for automatic detection of Mycobacterium tuberculosis

Microscopy images must be acquired at the optimal focal plane for the objects of interest in a scene. Although manual focusing is a standard task for a trained observer, automatic systems often fail to properly find the focal plane under different microscope imaging modalities such as bright field microscopy or phase contrast microscopy. This article assesses several autofocus algorithms applied in the study of fluorescence-labeled tuberculosis bacteria. The goal of this work was to find the optimal algorithm in order to build an automatic real-time system for diagnosing sputum smear samples, where both accuracy and computational time are important. We analyzed 13 focusing methods, ranging from well-known algorithms to the most recently proposed functions. We took into consideration criteria that are inherent to the autofocus function, such as accuracy, computational cost, and robustness to noise and to illumination changes. We also analyzed the additional benefit provided by preprocessing techniques based on morphological operators and image projection profiling.

An interacting multiple model filter-based autofocus strategy for confocal time-lapse microscopy

Gene expression and other cellular processes are stochastic, thus their study requires observing multiple events in multiple cells. Therefore, confocal microscopy cell imaging has recently gained much interest. In time-lapse imaging, adjustments are needed at short intervals to compensate for focus drift. There are several automated methods for this purpose. In general, before acquiring higher resolution images, software-based autofocus algorithms require a set of low-resolution images along the z-axis to determine the plane for which a predefined focusing function is maximized. These algorithms require 10-100 z-slices each time, and there is no fixed number or upper limit of required z-slices that ensures optimal focusing. The higher is this number, the stronger is photo bleaching, hampering the feasibility of long-time series measurements. We propose a new focusing strategy in time-lapse imaging. The algorithm relies on the nature and predictability of the focus drift. We first show that the focus drift curve is predictable within a small error bound in standard experimental setups. We, then, exploit the interacting multiple model filter algorithm to predict the drift at time, t, based on the measurement at time t-1. This allows a drastic reduction of the number of required z-slices for focus drift correction, largely overcoming the problem of photo bleaching. In addition, we propose a new set of functions for focusing in time-lapse imaging, derived from preexisting ones. We demonstrate the method's efficiency in time-lapse imaging of Escherichia coli cells expressing MS2d-GFP tagged RNA molecules.

Development of an automated MODS plate reader to detect early growth of Mycobacterium tuberculosis

In this work, an automated microscopic observation drug susceptibility (MODS) plate reader has been developed. The reader automatically handles MODS plates and after autofocussing digital images are acquired of the characteristic microscopic cording structures of Mycobacterium tuberculosis, which are the identification method utilized in the MODS technique to detect tuberculosis and multidrug resistant tuberculosis. In conventional MODS, trained technicians manually move the MODS plate on the stage of an inverted microscope while trying to locate and focus upon the characteristic microscopic cording colonies. In centres with high tuberculosis diagnostic demand, sufficient time may not be available to adequately examine all cultures. An automated reader would reduce labour time and the handling of M. tuberculosis cultures by laboratory personnel. Two hundred MODS culture images (100 from tuberculosis positive and 100 from tuberculosis negative sputum samples confirmed by a standard MODS reading using a commercial microscope) were acquired randomly using the automated MODS plate reader. A specialist analysed these digital images with the help of a personal computer and designated them as M. tuberculosis present or absent. The specialist considered four images insufficiently clear to permit a definitive reading. The readings from the 196 valid images resulted in a 100% agreement with the conventional nonautomated standard reading. The automated MODS plate reader combined with open-source MODS pattern recognition software provides a novel platform for high throughput automated tuberculosis diagnosis.