Artificial neural network classifier for the diagnosis of Parkinson's disease using [99mTc]TRODAT-1 and SPECT

Automated identification of uncertain cases in deep learning-based classification of dopamine transporter SPECT to improve clinical utility and acceptance

PURPOSE

Deep convolutional neural networks (CNN) are promising for automatic classification of dopamine transporter (DAT)-SPECT images. Reporting the certainty of CNN-based decisions is highly desired to flag cases that might be misclassified and, therefore, require particularly careful inspection by the user. The aim of the current study was to design and validate a CNN-based system for the identification of uncertain cases.

METHODS

A network ensemble (NE) combining five CNNs was trained for binary classification of [123I]FP-CIT DAT-SPECT images as "normal" or "neurodegeneration-typical reduction" with high accuracy (NE for classification, NEfC). An uncertainty detection module (UDM) was obtained by combining two additional NE, one trained for detection of "reduced" DAT-SPECT with high sensitivity, the other with high specificity. A case was considered "uncertain" if the "high sensitivity" NE and the "high specificity" NE disagreed. An internal "development" dataset of 1740 clinical DAT-SPECT images was used for training (n = 1250) and testing (n = 490). Two independent datasets with different image characteristics were used for testing only (n = 640, 645). Three established approaches for uncertainty detection were used for comparison (sigmoid, dropout, model averaging).

RESULTS

In the test data from the development dataset, the NEfC achieved 98.0% accuracy. 4.3% of all test cases were flagged as "uncertain" by the UDM: 2.5% of the correctly classified cases and 90% of the misclassified cases. NEfC accuracy among "certain" cases was 99.8%. The three comparison methods were less effective in labelling misclassified cases as "uncertain" (40-80%). These findings were confirmed in both additional test datasets.

CONCLUSION

The UDM allows reliable identification of uncertain [123I]FP-CIT SPECT with high risk of misclassification. We recommend that automatic classification of [123I]FP-CIT SPECT images is combined with an UDM to improve clinical utility and acceptance. The proposed UDM method ("high sensitivity versus high specificity") might be useful also for DAT imaging with other ligands and for other binary classification tasks.

From Recognition to Remedy: The Significance of Biomarkers in Neurodegenerative Disease Pathology

With the inexorable aging of the global populace, neurodegenerative diseases (NDs) like Alzheimer's disease (AD), Parkinson's disease (PD), and amyotrophic lateral sclerosis (ALS) pose escalating challenges, which are underscored by their socioeconomic repercussions. A pivotal aspect in addressing these challenges lies in the elucidation and application of biomarkers for timely diagnosis, vigilant monitoring, and effective treatment modalities. This review delineates the quintessence of biomarkers in the realm of NDs, elucidating various classifications and their indispensable roles. Particularly, the quest for novel biomarkers in AD, transcending traditional markers in PD, and the frontier of biomarker research in ALS are scrutinized. Emergent susceptibility and trait markers herald a new era of personalized medicine, promising enhanced treatment initiation especially in cases of SOD1-ALS. The discourse extends to diagnostic and state markers, revolutionizing early detection and monitoring, alongside progression markers that unveil the trajectory of NDs, propelling forward the potential for tailored interventions. The synergy between burgeoning technologies and innovative techniques like -omics, histologic assessments, and imaging is spotlighted, underscoring their pivotal roles in biomarker discovery. Reflecting on the progress hitherto, the review underscores the exigent need for multidisciplinary collaborations to surmount the challenges ahead, accelerate biomarker discovery, and herald a new epoch of understanding and managing NDs. Through a panoramic lens, this article endeavors to provide a comprehensive insight into the burgeoning field of biomarkers in NDs, spotlighting the promise they hold in transforming the diagnostic landscape, enhancing disease management, and illuminating the pathway toward efficacious therapeutic interventions.

Classification of Parkinson's Disease in Patch-Based MRI of Substantia Nigra

Parkinson's disease (PD) is a chronic and progressive neurological disease that mostly shakes and compromises the motor system of the human brain. Patients with PD can face resting tremors, loss of balance, bradykinesia, and rigidity problems. Complex patterns of PD, i.e., with relevance to other neurological diseases and minor changes in brain structure, make the diagnosis of this disease a challenge and cause inaccuracy of about 25% in the diagnostics. The research community utilizes different machine learning techniques for diagnosis using handcrafted features. This paper proposes a computer-aided diagnostic system using a convolutional neural network (CNN) to diagnose PD. CNN is one of the most suitable models to extract and learn the essential features of a problem. The dataset is obtained from Parkinson's Progression Markers Initiative (PPMI), which provides different datasets (benchmarks), such as T2-weighted MRI for PD and other healthy controls (HC). The mid slices are collected from each MRI. Further, these slices are registered for alignment. Since the PD can be found in substantia nigra (i.e., the midbrain), the midbrain region of the registered T2-weighted MRI slice is selected using the freehand region of interest technique with a 33 × 33 sized window. Several experiments have been carried out to ensure the validity of the CNN. The standard measures, such as accuracy, sensitivity, specificity, and area under the curve, are used to evaluate the proposed system. The evaluation results show that CNN provides better accuracy than machine learning techniques, such as naive Bayes, decision tree, support vector machine, and artificial neural network.

A fully automated approach involving neuroimaging and deep learning for Parkinson's disease detection and severity prediction

Three-dimensional magnetic resonance imaging has been proved to detect and predict the severity of progressive neurodegenerative disorders such as Parkinson's disease. The application of pre-processing with neuroimaging methods plays a vital role in post-processing for these problems. The development of technology over the years has enabled the use of deep learning methods such as convolutional neural networks (CNN) on magnetic resonance imaging (MRI) . In this study, the detection of Parkinson's disease and the prediction of disease severity were studied with 2D and 3D CNN using T1-weighted MRIs that were pre-processed with FLIRT image registration and BET non-brain tissue scraper. For 2D CNN, the median slices of the MR images in the sagittal, coronal, and axial planes were used separately and in combination. In addition, the whole brain for 3D CNN has been downsized. Considering the performance of the proposed methods, the highest results achieved for detecting Parkinson's disease were measured as 0.9620, 0.9452, 0.9407, and 0.9536 for Accuracy, F1 score, precision, and Recall, respectively. The highest result achieved for estimating the severity of Parkinson's disease was that 3D CNN was fed three times with a downsized whole MRI, which were measured for R, and R2 as 0.9150 and 0.8372, respectively. When the results obtained with the methods suggested within the scope of the study were examined, it was observed that the applied methods yielded promising performance.

Mining imaging and clinical data with machine learning approaches for the diagnosis and early detection of Parkinson's disease

Parkinson’s disease (PD) is a common, progressive, and currently incurable neurodegenerative movement disorder. The diagnosis of PD is challenging, especially in the differential diagnosis of parkinsonism and in early PD detection. Due to the advantages of machine learning such as learning complex data patterns and making inferences for individuals, machine-learning techniques have been increasingly applied to the diagnosis of PD, and have shown some promising results. Machine-learning-based imaging applications have made it possible to help differentiate parkinsonism and detect PD at early stages automatically in a number of neuroimaging studies. Comparative studies have shown that machine-learning-based SPECT image analysis applications in PD have outperformed conventional semi-quantitative analysis in detecting PD-associated dopaminergic degeneration, performed comparably well as experts’ visual inspection, and helped improve PD diagnostic accuracy of radiologists. Using combined multi-modal (imaging and clinical) data in these applications may further enhance PD diagnosis and early detection. To integrate machine-learning-based diagnostic applications into clinical systems, further validation and optimization of these applications are needed to make them accurate and reliable. It is anticipated that machine-learning techniques will further help improve differential diagnosis of parkinsonism and early detection of PD, which may reduce the error rate of PD diagnosis and help detect PD at pre-motor stage to make it possible for early treatments (e.g., neuroprotective treatment) to slow down PD progression, prevent severe motor symptoms from emerging, and relieve patients from suffering.

Artificial Intelligence for Brain Molecular Imaging

AI has been applied to brain molecular imaging for over 30 years. The past two decades, have seen explosive progress. AI applications span from operations processes such as attenuation correction and image generation, to disease diagnosis and prediction. As sophistication in AI software platforms increases, and the availability of large imaging data repositories become common, future studies will incorporate more multidimensional datasets and information that may truly reach "superhuman" levels in the field of brain imaging. However, even with a growing level of complexity, these advanced networks will still require human supervision for appropriate application and interpretation in medical practice.

Explainable AI to improve acceptance of convolutional neural networks for automatic classification of dopamine transporter SPECT in the diagnosis of clinically uncertain parkinsonian syndromes

Purpose

Deep convolutional neural networks (CNN) provide high accuracy for automatic classification of dopamine transporter (DAT) SPECT images. However, CNN are inherently black-box in nature lacking any kind of explanation for their decisions. This limits their acceptance for clinical use. This study tested layer-wise relevance propagation (LRP) to explain CNN-based classification of DAT-SPECT in patients with clinically uncertain parkinsonian syndromes.

Methods

The study retrospectively included 1296 clinical DAT-SPECT with visual binary interpretation as “normal” or “reduced” by two experienced readers as standard-of-truth. A custom-made CNN was trained with 1008 randomly selected DAT-SPECT. The remaining 288 DAT-SPECT were used to assess classification performance of the CNN and to test LRP for explanation of the CNN-based classification.

Results

Overall accuracy, sensitivity, and specificity of the CNN were 95.8%, 92.8%, and 98.7%, respectively. LRP provided relevance maps that were easy to interpret in each individual DAT-SPECT. In particular, the putamen in the hemisphere most affected by nigrostriatal degeneration was the most relevant brain region for CNN-based classification in all reduced DAT-SPECT. Some misclassified DAT-SPECT showed an “inconsistent” relevance map more typical for the true class label.

Conclusion

LRP is useful to provide explanation of CNN-based decisions in individual DAT-SPECT and, therefore, can be recommended to support CNN-based classification of DAT-SPECT in clinical routine. Total computation time of 3 s is compatible with busy clinical workflow. The utility of “inconsistent” relevance maps to identify misclassified cases requires further investigation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00259-021-05569-9.

Artificial intelligence techniques support nuclear medicine modalities to improve the diagnosis of Parkinson’s disease and Parkinsonian syndromes

Purpose

The aim of this review is to discuss the most significant contributions about the role of Artificial Intelligence (AI) techniques to support the diagnosis of movement disorders through nuclear medicine modalities.

Methods

The work is based on a selection of papers available on PubMed, Scopus and Web of Sciences. Articles not written in English were not considered in this study.

Results

Many papers are available concerning the increasing contribution of machine learning techniques to classify Parkinson’s disease (PD), Parkinsonian syndromes and Essential Tremor (ET) using data derived from brain SPECT with dopamine transporter radiopharmaceuticals. Other papers investigate by AI techniques data obtained by 123I-MIBG myocardial scintigraphy to differentially diagnose PD and other Parkinsonian syndromes.

Conclusion

The recent literature provides strong evidence that AI techniques can play a fundamental role in the diagnosis of movement disorders by means of nuclear medicine modalities, therefore paving the way towards personalized medicine.

Towards clinical application of image mining: a systematic review on artificial intelligence and radiomics

PURPOSE

The aim of this systematic review was to analyse literature on artificial intelligence (AI) and radiomics, including all medical imaging modalities, for oncological and non-oncological applications, in order to assess how far the image mining research stands from routine medical application. To do this, we applied a trial phases classification inspired from the drug development process.

METHODS

Among the articles we considered for inclusion from PubMed were multimodality AI and radiomics investigations, with a validation analysis aimed at relevant clinical objectives. Quality assessment of selected papers was performed according to the QUADAS-2 criteria. We developed the phases classification criteria for image mining studies.

RESULTS

Overall 34,626 articles were retrieved, 300 were selected applying the inclusion/exclusion criteria, and 171 high-quality papers (QUADAS-2 ≥ 7) were identified and analysed. In 27/171 (16%), 141/171 (82%), and 3/171 (2%) studies the development of an AI-based algorithm, radiomics model, and a combined radiomics/AI approach, respectively, was described. A total of 26/27(96%) and 1/27 (4%) AI studies were classified as phase II and III, respectively. Consequently, 13/141 (9%), 10/141 (7%), 111/141 (79%), and 7/141 (5%) radiomics studies were classified as phase 0, I, II, and III, respectively. All three radiomics/AI studies were categorised as phase II trials.

CONCLUSIONS

The results of the studies are promising but still not mature enough for image mining tools to be implemented in the clinical setting and be widely used. The transfer learning from the well-known drug development process, with some specific adaptations to the image mining discipline could represent the most effective way for radiomics and AI algorithms to become the standard of care tools.

Discriminative analysis of Parkinson's disease based on whole-brain functional connectivity

Recently, there has been an increasing emphasis on applications of pattern recognition and neuroimaging techniques in the effective and accurate diagnosis of psychiatric or neurological disorders. In the present study, we investigated the whole-brain resting-state functional connectivity patterns of Parkinson's disease (PD), which are expected to provide additional information for the clinical diagnosis and treatment of this disease. First, we computed the functional connectivity between each pair of 116 regions of interest derived from a prior atlas. The most discriminative features based on Kendall tau correlation coefficient were then selected. A support vector machine classifier was employed to classify 21 PD patients with 26 demographically matched healthy controls. This method achieved a classification accuracy of 93.62% using leave-one-out cross-validation, with a sensitivity of 90.47% and a specificity of 96.15%. The majority of the most discriminative functional connections were located within or across the default mode, cingulo-opercular and frontal-parietal networks and the cerebellum. These disease-related resting-state network alterations might play important roles in the pathophysiology of this disease. Our results suggest that analyses of whole-brain resting-state functional connectivity patterns have the potential to improve the clinical diagnosis and treatment evaluation of PD.

Optimization of imaging parameters for SPECT scans of [99mTc]TRODAT-1 using Taguchi analysis

Parkinson's disease (PD) is a neurodegenerative disease characterized by progressive loss of dopaminergic neurons in the basal ganglia. Single photon emission computed tomography (SPECT) scans using [99mTc]TRODAT-1 can image dopamine transporters and provide valuable diagnostic information of PD. In this study, we optimized the scanning parameters for [99mTc]TRODAT-1/SPECT using the Taguchi analysis to improve image quality. SPECT scans were performed on forty-five healthy volunteers according to an L9 orthogonal array. Three parameters were considered, including the injection activity, uptake duration, and acquisition time per projection. The signal-to-noise ratio (SNR) was calculated from the striatum/occipital activity ratio as an image quality index. Ten healthy subjects and fifteen PD patients were used to verify the optimal parameters. The estimated optimal parameters were 962 MBq for [99mTc]TRODAT-1 injection, 260 min for uptake duration, and 60 s/projection for data acquisition. The uptake duration and time per projection were the two dominant factors which had an F-value of 18.638 (38%) and 25.933 (53%), respectively. Strong cross interactions existed between the injection activity/uptake duration and injection activity/time per projection. Therefore, under the consideration of as low as reasonably achievable (ALARA) for radiation protection, we can decrease the injection activity to 740 MBq. The image quality remains almost the same for clinical applications.

Automatic classification of early Parkinson's disease with multi-modal MR imaging

Background

In recent years, neuroimaging has been increasingly used as an objective method for the diagnosis of Parkinson's disease (PD). Most previous studies were based on invasive imaging modalities or on a single modality which was not an ideal diagnostic tool. In this study, we developed a non-invasive technology intended for use in the diagnosis of early PD by integrating the advantages of various modals.

Materials and Methods

Nineteen early PD patients and twenty-seven normal volunteers participated in this study. For each subject, we collected resting-state functional magnetic resonance imaging (rsfMRI) and structural images. For the rsfMRI images, we extracted the characteristics at three different levels: ALFF (amplitude of low-frequency fluctuations), ReHo (regional homogeneity) and RFCS (regional functional connectivity strength). For the structural images, we extracted the volume characteristics from the gray matter (GM), the white matter (WM) and the cerebrospinal fluid (CSF). A two-sample t-test was used for the feature selection, and then the remaining features were fused for classification. Finally a classifier for early PD patients and normal control subjects was identified from support vector machine training. The performance of the classifier was evaluated using the leave-one-out cross-validation method.

Results

Using the proposed methods to classify the data set, good results (accuracy  = 86.96%, sensitivity  = 78.95%, specificity  = 92.59%) were obtained.

Conclusions

This method demonstrates a promising diagnosis performance by the integration of information from a variety of imaging modalities, and it shows potential for improving the clinical diagnosis and treatment of PD.

Comparison of two neural network classifiers in the differential diagnosis of essential tremor and Parkinson's disease by (123)I-FP-CIT brain SPECT

PURPOSE

To contribute to the differentiation of Parkinson's disease (PD) and essential tremor (ET), we compared two different artificial neural network classifiers using (123)I-FP-CIT SPECT data, a probabilistic neural network (PNN) and a classification tree (ClT).

METHODS

(123)I-FP-CIT brain SPECT with semiquantitative analysis was performed in 216 patients: 89 with ET, 64 with PD with a Hoehn and Yahr (H&Y) score of ≤2 (early PD), and 63 with PD with a H&Y score of ≥2.5 (advanced PD). For each of the 1,000 experiments carried out, 108 patients were randomly selected as the PNN training set, while the remaining 108 validated the trained PNN, and the percentage of the validation data correctly classified in the three groups of patients was computed. The expected performance of an "average performance PNN" was evaluated. In analogy, for ClT 1,000 classification trees with similar structures were generated.

RESULTS

For PNN, the probability of correct classification in patients with early PD was 81.9±8.1% (mean±SD), in patients with advanced PD 78.9±8.1%, and in ET patients 96.6±2.6%. For ClT, the first decision rule gave a mean value for the putamen of 5.99, which resulted in a probability of correct classification of 93.5±3.4%. This means that patients with putamen values >5.99 were classified as having ET, while patients with putamen values <5.99 were classified as having PD. Furthermore, if the caudate nucleus value was higher than 6.97 patients were classified as having early PD (probability 69.8±5.3%), and if the value was <6.97 patients were classified as having advanced PD (probability 88.1%±8.8%).

CONCLUSION

These results confirm that PNN achieved valid classification results. Furthermore, ClT provided reliable cut-off values able to differentiate ET and PD of different severities.

Differential diagnosis of parkinsonian syndromes using PCA-based functional imaging features

In the current paper, we describe methodologies for single subject differential diagnosis of degenerative brain disorders using multivariate principal component analysis (PCA) of functional imaging scans. An automated routine utilizing these methods is applied to positron emission tomography (PET) brain data to distinguish several discrete parkinsonian movement disorders with similar clinical manifestations. Disease specific expressions of voxel-based spatial covariance patterns are predetermined using the Scaled Subprofile Model (SSM/PCA) and a scalar measure of the manifestation of each pattern in prospective subject images is subsequently derived. Scores are automatically compared to reference values generated for each pathological condition in a corresponding set of patient and control scans. Diagnostic outcome is optimized using strategies such as the derivation of patterns in a voxel subspace that reflects contrasting image characteristics between conditions, or by using an independent patient population as controls. The prediction models for two, three and four way classification problems using direct scalar comparison as well as classical discriminant analysis are assessed in a composite training population comprised of three different patient classes and normal controls, and validated in a similar independent test population. Results illustrate that highly accurate diagnosis can often be achieved by simple comparison of scores utilizing optimized patterns.

Multispectral imaging and artificial neural network: mimicking the management decision of the clinician facing pigmented skin lesions

Various instruments based on acquisition and elaboration of images of pigmented skin lesions have been developed in an attempt to in vivo establish whether a lesion is a melanoma or not. Although encouraging, the response of these instruments, e.g. epiluminescence microscopy, reflectance spectrophotometry and fluorescence imaging, cannot currently replace the well-established diagnostic procedures. However, in place of the approach to instrumentally assess the diagnosis of the lesion, recent studies suggest that instruments should rather reproduce the assessment by an expert clinician of whether a lesion has to be excised or not. The aim of this study was to evaluate the performance of a spectrophotometric system to mimic such a decision. The study involved 1794 consecutively recruited patients with 1966 doubtful cutaneous pigmented lesions excised for histopathological diagnosis and 348 patients with 1940 non-excised lesions because clinically reassuring. Images of all these lesions were acquired in vivo with a multispectral imaging system. The data set was randomly divided into a train (802 reassuring and 1003 excision-needing lesions, including 139 melanomas), a verify (464 reassuring and 439 excision-needing lesions, including 72 melanomas) and a test set (674 reassuring and 524 excision-needing lesions, including 76 melanomas). An artificial neural network (ANN(1)) was set up to perform the classification of the lesions as excision-needing or reassuring, according to the expert clinicians' decision on how to manage each examined lesion. In the independent test set, the system was able to emulate the clinicians with a sensitivity of 88% and a specificity of 80%. Of the 462 correctly classified as excision-needing lesions, 72 (95%) were melanomas. No major variations in receiver operating characteristic curves were found between the test and the train/verify sets. On the same data set, a further artificial neural network (ANN(2)) was then architected to perform classification of the lesions as melanoma or non-melanoma, according to the histological diagnosis. Having set the sensitivity in recognizing melanoma to 95%, ANN(1) resulted to be significantly better in the classification of reassuring lesions than ANN(2). This study suggests that multispectral image analysis and artificial neural networks could be used to support primary care physicians or general practitioners in identifying pigmented skin lesions that require further investigations.