LSTM-CNN: An efficient diagnostic network for Parkinson's disease utilizing dynamic handwriting analysis

BACKGROUND AND OBJECTIVES

Dynamic handwriting analysis, due to its noninvasive and readily accessible nature, has emerged as a vital adjunctive method for the early diagnosis of Parkinson's disease (PD). An essential step involves analysing subtle variations in signals to quantify PD dysgraphia. Although previous studies have explored extracting features from the overall signal, they may ignore the potential importance of local signal segments. In this study, we propose a lightweight network architecture to analyse dynamic handwriting signal segments of patients and present visual diagnostic results, providing an efficient diagnostic method.

METHODS

To analyse subtle variations in handwriting, we investigate time-dependent patterns in local representation of handwriting signals. Specifically, we segment the handwriting signal into fixed-length sequential segments and design a compact one-dimensional (1D) hybrid network to extract discriminative temporal features for classifying each local segment. Finally, the category of the handwriting signal is fully diagnosed through a majority voting scheme.

RESULTS

The proposed method achieves impressive diagnostic performance on the new DraWritePD dataset (with an accuracy of 96.2%, sensitivity of 94.5% and specificity of 97.3%) and the well-established PaHaW dataset (with an accuracy of 90.7%, sensitivity of 94.3% and specificity of 87.5%). Moreover, the network architecture stands out for its excellent lightweight design, occupying a mere 0.084M parameters, with only 0.59M floating-point operations. It also exhibits nearly real-time CPU inference performance, with the inference time for a single handwriting signal ranging from 0.106 to 0.220 s.

CONCLUSIONS

We present a series of experiments with extensive analysis, which systematically demonstrate the effectiveness and efficiency of the proposed method in quantifying dysgraphia for a precise diagnosis of PD.

Determination of esophageal squamous cell carcinoma and gastric adenocarcinoma on raw tissue using Raman spectroscopy

BACKGROUND

Cancer detection is a global research focus, and novel, rapid, and label-free techniques are being developed for routine clinical practice. This has led to the development of new tools and techniques from the bench side to routine clinical practice. In this study, we present a method that uses Raman spectroscopy (RS) to detect cancer in unstained formalin-fixed, resected specimens of the esophagus and stomach. Our method can record a clear Raman-scattered light spectrum in these specimens, confirming that the Raman-scattered light spectrum changes because of the histological differences in the mucosal tissue.

AIM

To evaluate the use of Raman-scattered light spectrum for detecting endoscop-ically resected specimens of esophageal squamous cell carcinoma (SCC) and gastric adenocarcinoma (AC).

METHODS

We created a Raman device that is suitable for observing living tissues, and attempted to acquire Raman-scattered light spectra in endoscopically resected specimens of six esophageal tissues and 12 gastric tissues. We evaluated formalin-fixed tissues using this technique and captured shifts at multiple locations based on feasibility, ranging from six to 19 locations 200 microns apart in the vertical and horizontal directions. Furthermore, a correlation between the obtained Raman scattered light spectra and histopathological diagnosis was performed.

RESULTS

We successfully obtained Raman scattered light spectra from all six esophageal and 12 gastric specimens. After data capture, the tissue specimens were sent for histopathological analysis for further processing because RS is a label-free methodology that does not cause tissue destruction or alterations. Based on data analysis of molecular-level substrates, we established cut-off values for the diagnosis of esophageal SCC and gastric AC. By analyzing specific Raman shifts, we developed an algorithm to identify the range of esophageal SCC and gastric AC with an accuracy close to that of histopathological diagnoses.

CONCLUSION

Our technique provides qualitative information for real-time morphological diagnosis. However, further in vivo evaluations require an excitation light source with low human toxicity and large amounts of data for validation.

MFB-LANN: A lightweight and updatable myocardial infarction diagnosis system based on convolutional neural networks and active learning

BACKGROUND AND OBJECTIVES

12 leads electrocardiogram (ECG) are widely used to diagnose myocardial infarction (MI). Generally, the symptoms of MI can be reflected by waveforms in the heartbeat, and the contribution of different ECG leads to different types of MI is different. Therefore, it is significant to use the heartbeat waveform features and the lead relationship features for multi-category MI diagnosis. Moreover, the challenge of individual differences and lightweight algorithms also need to be further resolved and explored in the ECG automatic diagnosis system.

METHODS

This paper presents a lightweight MI diagnosis system named multi-feature-branch lead attention neural network (MFB-LANN) via 12 leads ECG signals. It is designed based on the characteristics of the ECG lead. Specifically, 12 independent feature branches correspond to different leads, and each branch contains different convolutional layers to extract features in the heartbeat, then a novel attention module is developed named lead attention mechanism (LAM) to assign different weights to each feature branch. Finally all the weighted feature branches are fused for classification. Furthermore, to overcome individual differences, patient-specific scheme and active learning (AL) are used to train and update the model iteratively.

RESULTS

Experimental results based on Physikalisch-Technische Bundesanstalt (PTB) database shows that the MFB-LANN achieved satisfactory results with accuracy of 99.63% based on 5-fold cross validation under the intra-patient scheme. The patient-specific experiment yielded an average accuracy of 96.99% compared to the state-of-the-art. By contrast, the model achieved acceptable results on the hybrid database (PTB and PTB-XL), especially achieving 94.19% accuracy after the update. Moreover, the system can complete the update process and real-time diagnosis on the ARM Cortex-A72 platform.

CONCLUSIONS

Experiments show that the proposed method for MI diagnosis has more obvious advantages compared to other recent methods, and it has great potential to be applied to the mobile medical field.

Noninvasive assessment of oral lesions using elastic light single-scattering spectroscopy: a pilot study

PURPOSE

In the present study, we investigated the potential application of elastic light single-scattering spectroscopy (ELSSS) as a noninvasive, adjunctive tool to differentiate between malignant and benign oral lesions in vivo.

METHODS

ELSSS spectra were acquired from 52 oral lesions of 47 patients prior to surgical biopsy using a single optical fiber probe. The sign of the spectral slope was used as a diagnostic parameter and was compared to the histopathology findings to obtain sensitivity and specificity of the ELSSS system in differentiating between benign and malignant tissues.

RESULTS

The sign of the spectral slope was positive for the benign tissues and negative for the malignant tissues. Nine malignant lesions and one high-grade dysplasia were correctly classified as cancerous. Six out of the ten low-grade dysplasia were correctly classified as cancerous, and four of them were misclassified as benign. Thirty benign lesions were correctly classified as benign, and two were misclassified as malignant. Our results indicate that the sign of the spectral slope enables the differentiation between malignant and benign oral lesions with a sensitivity of 80% and specificity of 94%.

CONCLUSIONS

ELSSS has the potential to be developed as an adjunctive screening tool in the noninvasive evaluation of oral lesions in vivo. This new diagnostic system may reduce the number of unnecessary biopsies.

Preclinical ex vivo evaluation of the diagnostic performance of a new device for in situ label-free fluorescence spectral analysis of breast masses

OBJECTIVES

To assess the diagnostic performance of a new device for in situ label-free fluorescence spectral analysis of breast masses in freshly removed surgical specimens, in preparation for its clinical development.

METHODS

Sixty-four breast masses from consenting patients who had undergone either a lumpectomy or a mastectomy were included. Label-free fluorescence spectral acquisitions were obtained with a 25G fibre-containing needle inserted into the mass. Data from benign and malignant masses were compared to establish the most discriminating thresholds and measurement algorithms. Accuracy was verified using the bootstrap method.

RESULTS

The final histological examination revealed 44 invasive carcinomas and 20 benign lesions. The maximum intensity of fluorescence signal was discriminant between benign and malignant masses (p < .0001) whatever their sizes. Statistical analysis indicated that choosing five random measurements per mass was the best compromise to obtain high sensitivity and high negative predictive value with the fewest measurements. Thus, malignant tumours were identified with a mean sensitivity, specificity, negative and positive predictive value of 98.8%, 85.4%, 97.2% and 93.5%, respectively.

CONCLUSION

This new in situ tissue autofluorescence evaluation device allows accurate discrimination between benign and malignant breast masses and deserves clinical development.

KEY POINTS

• A new device allows in situ label-free fluorescence analysis of ex vivo breast masses • Maximum fluorescence intensity discriminates benign from malignant masses (p < .0001) • Five random measurements allow a high negative predictive value (97.2%).

Raman Spectroscopy for the Endoscopic Diagnosis of Esophageal, Gastric, and Colonic Diseases

Globally white-light endoscopy with biopsy sampling is the gold standard diagnostic modality for esophageal, gastric, and colonic pathologies. However, there is overwhelming evidence to highlight the deficiencies of an approach based predominantly on eyeball visualization. Biopsy sampling is also problematic due in part to excessive sampling and hence attendant cost. Various innovations are currently taking place in the endoscopic domain to aid operators in diagnosis forming. These include narrow band imaging which aims to enhance the surface anatomy and vasculature, and confocal laser endomicroscopy which provides real time histological information. However, both of these tools are limited by the skill of the operator and the extensive learning curve associated with their use. There is a gap therefore for a new form of technology that relies solely on an objective measure of disease and reduces the need for biopsy sampling. Raman spectroscopy (RS) is a potential platform that aims to satisfy these criteria. It enables a fingerprint capture of tissue in relation to the protein, DNA, and lipid content. This focused review highlights the strong potential for the use of RS during endoscopic gastroenterological examination.