Edge Artificial Intelligence (AI) for real-time automatic quantification of filariasis in mobile microscopy

Filariasis, a neglected tropical disease caused by roundworms, is a significant public health concern in many tropical countries. Microscopic examination of blood samples can detect and differentiate parasite species, but it is time consuming and requires expert microscopists, a resource that is not always available. In this context, artificial intelligence (AI) can assist in the diagnosis of this disease by automatically detecting and differentiating microfilariae. In line with the target product profile for lymphatic filariasis as defined by the World Health Organization, we developed an edge AI system running on a smartphone whose camera is aligned with the ocular of an optical microscope that detects and differentiates filarias species in real time without the internet connection. Our object detection algorithm that uses the Single-Shot Detection (SSD) MobileNet V2 detection model was developed with 115 cases, 85 cases with 1903 fields of view and 3342 labels for model training, and 30 cases with 484 fields of view and 873 labels for model validation before clinical validation, is able to detect microfilariae at 10x magnification and distinguishes four species of them at 40x magnification: Loa loa, Mansonella perstans, Wuchereria bancrofti, and Brugia malayi. We validated our augmented microscopy system in the clinical environment by replicating the diagnostic workflow encompassed examinations at 10x and 40x with the assistance of the AI models analyzing 18 samples with the AI running on a middle range smartphone. It achieved an overall precision of 94.14%, recall of 91.90% and F1 score of 93.01% for the screening algorithm and 95.46%, 97.81% and 96.62% for the species differentiation algorithm respectively. This innovative solution has the potential to support filariasis diagnosis and monitoring, particularly in resource-limited settings where access to expert technicians and laboratory equipment is scarce.

Digital Microscopy Augmented by Artificial Intelligence to Interpret Bone Marrow Samples for Hematological Diseases

Analysis of bone marrow aspirates (BMAs) is an essential step in the diagnosis of hematological disorders. This analysis is usually performed based on a visual examination of samples under a conventional optical microscope, which involves a labor-intensive process, limited by clinical experience and subject to high observer variability. In this work, we present a comprehensive digital microscopy system that enables BMA analysis for cell type counting and differentiation in an efficient and objective manner. This system not only provides an accessible and simple method to digitize, store, and analyze BMA samples remotely but is also supported by an Artificial Intelligence (AI) pipeline that accelerates the differential cell counting process and reduces interobserver variability. It has been designed to integrate AI algorithms with the daily clinical routine and can be used in any regular hospital workflow.

Integration of longitudinal deep-radiomics and clinical data improves the prediction of durable benefits to anti-PD-1/PD-L1 immunotherapy in advanced NSCLC patients

BACKGROUND

Identifying predictive non-invasive biomarkers of immunotherapy response is crucial to avoid premature treatment interruptions or ineffective prolongation. Our aim was to develop a non-invasive biomarker for predicting immunotherapy clinical durable benefit, based on the integration of radiomics and clinical data monitored through early anti-PD-1/PD-L1 monoclonal antibodies treatment in patients with advanced non-small cell lung cancer (NSCLC).

METHODS

In this study, 264 patients with pathologically confirmed stage IV NSCLC treated with immunotherapy were retrospectively collected from two institutions. The cohort was randomly divided into a training (n = 221) and an independent test set (n = 43), ensuring the balanced availability of baseline and follow-up data for each patient. Clinical data corresponding to the start of treatment was retrieved from electronic patient records, and blood test variables after the first and third cycles of immunotherapy were also collected. Additionally, traditional radiomics and deep-radiomics features were extracted from the primary tumors of the computed tomography (CT) scans before treatment and during patient follow-up. Random Forest was used to implementing baseline and longitudinal models using clinical and radiomics data separately, and then an ensemble model was built integrating both sources of information.

RESULTS

The integration of longitudinal clinical and deep-radiomics data significantly improved clinical durable benefit prediction at 6 and 9 months after treatment in the independent test set, achieving an area under the receiver operating characteristic curve of 0.824 (95% CI: [0.658,0.953]) and 0.753 (95% CI: [0.549,0.931]). The Kaplan-Meier survival analysis showed that, for both endpoints, the signatures significantly stratified high- and low-risk patients (p-value< 0.05) and were significantly correlated with progression-free survival (PFS6 model: C-index 0.723, p-value = 0.004; PFS9 model: C-index 0.685, p-value = 0.030) and overall survival (PFS6 models: C-index 0.768, p-value = 0.002; PFS9 model: C-index 0.736, p-value = 0.023).

CONCLUSIONS

Integrating multidimensional and longitudinal data improved clinical durable benefit prediction to immunotherapy treatment of advanced non-small cell lung cancer patients. The selection of effective treatment and the appropriate evaluation of clinical benefit are important for better managing cancer patients with prolonged survival and preserving quality of life.

Deep learning-based lesion subtyping and prediction of clinical outcomes in COVID-19 pneumonia using chest CT

The main objective of this work is to develop and evaluate an artificial intelligence system based on deep learning capable of automatically identifying, quantifying, and characterizing COVID-19 pneumonia patterns in order to assess disease severity and predict clinical outcomes, and to compare the prediction performance with respect to human reader severity assessment and whole lung radiomics. We propose a deep learning based scheme to automatically segment the different lesion subtypes in nonenhanced CT scans. The automatic lesion quantification was used to predict clinical outcomes. The proposed technique has been independently tested in a multicentric cohort of 103 patients, retrospectively collected between March and July of 2020. Segmentation of lesion subtypes was evaluated using both overlapping (Dice) and distance-based (Hausdorff and average surface) metrics, while the proposed system to predict clinically relevant outcomes was assessed using the area under the curve (AUC). Additionally, other metrics including sensitivity, specificity, positive predictive value and negative predictive value were estimated. 95% confidence intervals were properly calculated. The agreement between the automatic estimate of parenchymal damage (%) and the radiologists' severity scoring was strong, with a Spearman correlation coefficient (R) of 0.83. The automatic quantification of lesion subtypes was able to predict patient mortality, admission to the Intensive Care Units (ICU) and need for mechanical ventilation with an AUC of 0.87, 0.73 and 0.68 respectively. The proposed artificial intelligence system enabled a better prediction of those clinically relevant outcomes when compared to the radiologists' interpretation and to whole lung radiomics. In conclusion, deep learning lesion subtyping in COVID-19 pneumonia from noncontrast chest CT enables quantitative assessment of disease severity and better prediction of clinical outcomes with respect to whole lung radiomics or radiologists' severity score.

Mobile microscopy and telemedicine platform assisted by deep learning for the quantification of Trichuris trichiura infection

Soil-transmitted helminths (STH) are the most prevalent pathogens among the group of neglected tropical diseases (NTDs). The Kato-Katz technique is the diagnosis method recommended by the World Health Organization (WHO) although it often presents a decreased sensitivity in low transmission settings and it is labour intensive. Visual reading of Kato-Katz preparations requires the samples to be analyzed in a short period of time since its preparation. Digitizing the samples could provide a solution which allows to store the samples in a digital database and perform remote analysis. Artificial intelligence (AI) methods based on digitized samples can support diagnosis by performing an objective and automatic quantification of disease infection. In this work, we propose an end-to-end pipeline for microscopy image digitization and automatic analysis of digitized images of STH. Our solution includes (a) a digitization system based on a mobile app that digitizes microscope samples using a 3D printed microscope adapter, (b) a telemedicine platform for remote analysis and labelling, and (c) novel deep learning algorithms for automatic assessment and quantification of parasitological infections by STH. The deep learning algorithm has been trained and tested on 51 slides of stool samples containing 949 Trichuris spp. eggs from 6 different subjects. The algorithm evaluation was performed using a cross-validation strategy, obtaining a mean precision of 98.44% and a mean recall of 80.94%. The results also proved the potential of generalization capability of the method at identifying different types of helminth eggs. Additionally, the AI-assisted quantification of STH based on digitized samples has been compared to the one performed using conventional microscopy, showing a good agreement between measurements. In conclusion, this work has presented a comprehensive pipeline using smartphone-assisted microscopy. It is integrated with a telemedicine platform for automatic image analysis and quantification of STH infection using AI models.

Classification of Interstitial Lung Abnormality Patterns with an Ensemble of Deep Convolutional Neural Networks

Subtle interstitial changes in the lung parenchyma of smokers, known as Interstitial Lung Abnormalities (ILA), have been associated with clinical outcomes, including mortality, even in the absence of Interstitial Lung Disease (ILD). Although several methods have been proposed for the automatic identification of more advanced Interstitial Lung Disease (ILD) patterns, few have tackled ILA, which likely precedes the development ILD in some cases. In this context, we propose a novel methodology for automated identification and classification of ILA patterns in computed tomography (CT) images. The proposed method is an ensemble of deep convolutional neural networks (CNNs) that detect more discriminative features by incorporating two, two-and-a-half and three- dimensional architectures, thereby enabling more accurate classification. This technique is implemented by first training each individual CNN, and then combining its output responses to form the overall ensemble output. To train and test the system we used 37424 radiographic tissue samples corresponding to eight different parenchymal feature classes from 208 CT scans. The resulting ensemble performance including an average sensitivity of 91,41% and average specificity of 98,18% suggests it is potentially a viable method to identify radiographic patterns that precede the development of ILD.

A SR-NET 3D-TO-2D ARCHITECTURE FOR PARASEPTAL EMPHYSEMA SEGMENTATION

Paraseptal emphysema (PSE) is a relatively unexplored emphysema subtype that is usually asymptomatic, but recently associated with interstitial lung abnormalities which are related with clinical outcomes, including mortality. Previous local-based methods for emphysema subtype quantification do not properly characterize PSE. This is in part for their inability to properly capture the global aspect of the disease, as some the PSE lesions can involved large regions along the chest wall. It is our assumption, that path-based approaches are not well-suited to identify this subtype and segmentation is a better paradigm. In this work we propose and introduce the Slice-Recovery network (SR-Net) that leverages 3D contextual information for 2D segmentation of PSE lesions in CT images. For that purpose, a novel convolutional network architecture is presented, which follows an encoding-decoding path that processes a 3D volume to generate a 2D segmentation map. The dataset used for training and testing the method comprised 664 images, coming from 111 CT scans. The results demonstrate the benefit of the proposed approach which incorporate 3D context information to the network and the ability of the proposed method to identify and segment PSE lesions with different sizes even in the presence of other emphysema subtypes in an advanced stage.

EMPHYSEMA CLASSIFICATION USING A MULTI-VIEW CONVOLUTIONAL NETWORK

In this article we propose and validate a fully automatic tool for emphysema classification in Computed Tomography (CT) images. We hypothesize that a relatively simple Convolutional Neural Network (CNN) architecture can learn even better discriminative features from the input data compared with more complex and deeper architectures. The proposed architecture is comprised of only 4 convolutional and 3 pooling layers, where the input corresponds to a 2.5D multiview representation of the pulmonary segment tissue to classify, corresponding to axial, sagittal and coronal views. The proposed architecture is compared to similar 2D CNN and 3D CNN, and to more complex architectures which involve a larger number of parameters (up to six times larger). This method has been evaluated in 1553 tissue samples, and achieves an overall sensitivity of 81.78 % and a specificity of 97.34%, and results show that the proposed method outperforms deeper state-of-the-art architectures particularly designed for lung pattern classification. The method shows satisfactory results in full-lung classification.