Smartphone-based detection of COVID-19 and associated pneumonia using thermal imaging and a transfer learning algorithm.

COVID-19-related pneumonia is typically diagnosed using chest x-ray or computed tomography images. However, these techniques can only be used in hospitals. In contrast, thermal cameras are portable, inexpensive devices that can be connected to smartphones. Thus, they can be used to detect and monitor medical conditions outside hospitals. Herein, a smartphone-based application using thermal images of a human back was developed for COVID-19 detection. Image analysis using a deep learning algorithm revealed a sensitivity and specificity of 88.7% and 92.3%, respectively. The findings support the future use of noninvasive thermal imaging in primary screening for COVID-19 and associated pneumonia.

Automated thermal imaging monitors the local response to cervical cancer brachytherapy.

Malignant tumors have high metabolic and perfusion rates, which result in a unique temperature distribution as compared to healthy tissues. Here, we sought to characterize the thermal response of the cervix following brachytherapy in women with advanced cervical carcinoma. Six patients underwent imaging with a thermal camera before a brachytherapy treatment session and after a 7-day follow-up period. A designated algorithm was used to calculate and store the texture parameters of the examined tissues across all time points. We used supervised machine learning classification methods (K Nearest Neighbors and Support Vector Machine) and unsupervised machine learning classification (K-means). Our algorithms demonstrated a 100% detection rate for physiological changes in cervical tumors before and after brachytherapy. Thus, we showed that thermal imaging combined with advanced feature extraction could potentially be used to detect tissue-specific changes in the cervix in response to local brachytherapy for cervical cancer.

Automated processing of thermal imaging to detect COVID-19.

Rapid and sensitive screening tools for SARS-CoV-2 infection are essential to limit the spread of COVID-19 and to properly allocate national resources. Here, we developed a new point-of-care, non-contact thermal imaging tool to detect COVID-19, based on advanced image processing algorithms. We captured thermal images of the backs of individuals with and without COVID-19 using a portable thermal camera that connects directly to smartphones. Our novel image processing algorithms automatically extracted multiple texture and shape features of the thermal images and achieved an area under the curve (AUC) of 0.85 in COVID-19 detection with up to 92% sensitivity. Thermal imaging scores were inversely correlated with clinical variables associated with COVID-19 disease progression. In summary, we show, for the first time, that a hand-held thermal imaging device can be used to detect COVID-19. Non-invasive thermal imaging could be used to screen for COVID-19 in out-of-hospital settings, especially in low-income regions with limited imaging resources.

Assessment of blood distribution in response to post-surgical steal syndrome: A novel technique based on Thermo-Anatomical Segmentation.

The distal ischemic steal syndrome (ISS) is a complication following the construction of an arteriovenous (A-V) access for hemodialysis. The ability to non-invasively monitor changes in skin microcirculation improves both the diagnosis and treatment of vascular diseases. In this study, we propose a novel technique for evaluating the palms' blood distribution following arteriovenous access, based on thermal imaging. Furthermore, we utilize the thermal images to identify typical recovery patterns of patients that underwent this surgery and show that thermal images taken post-surgery reflect the patient's follow-up status. Thermal photographs were taken by a portable thermal camera from both hands before and after the A-V access surgery, and one month following the surgery, from ten dialysis patients. A novel term "Thermo-Anatomical Segmentation", which enables a functional assessment of palm blood distribution was defined. Based on this segmentation it was shown that the greatest change after surgery was in the most distal region, the fingertips (p < 0.05). In addition, the changes in palm blood distribution in both hands were synchronized, which indicates a bilateral effect. An unsupervised machine learning model revealed two variables that determine the recovery pattern following the surgery: the palms' temperature difference pre- and post-surgery and the post-surgery difference between the treated and untreated hand. Our proposed framework provides a new technique for quantitative assessment of the palm's blood distribution. This technique may improve the clinical treatment of patients with vascular disease, particularly the patient-specific follow-up, in clinics as well as in homecare.

Automated thermal imaging for the detection of fatty liver disease.

Non-alcoholic fatty liver disease (NAFLD) comprises a spectrum of progressive liver pathologies, ranging from simple steatosis to non-alcoholic steatohepatitis (NASH), fibrosis and cirrhosis. A liver biopsy is currently required to stratify high-risk patients, and predicting the degree of liver inflammation and fibrosis using non-invasive tests remains challenging. Here, we sought to develop a novel, cost-effective screening tool for NAFLD based on thermal imaging. We used a commercially available and non-invasive thermal camera and developed a new image processing algorithm to automatically predict disease status in a small animal model of fatty liver disease. To induce liver steatosis and inflammation, we fed C57/black female mice (8 weeks old) a methionine-choline deficient diet (MCD diet) for 6 weeks. We evaluated structural and functional liver changes by serial ultrasound studies, histopathological analysis, blood tests for liver enzymes and lipids, and measured liver inflammatory cell infiltration by flow cytometry. We developed an image processing algorithm that measures relative spatial thermal variation across the skin covering the liver. Thermal parameters including temperature variance, homogeneity levels and other textural features were fed as input to a t-SNE dimensionality reduction algorithm followed by k-means clustering. During weeks 3,4, and 5 of the experiment, our algorithm demonstrated a 100% detection rate and classified all mice correctly according to their disease status. Direct thermal imaging of the liver confirmed the presence of changes in surface thermography in diseased livers. We conclude that non-invasive thermal imaging combined with advanced image processing and machine learning-based analysis successfully correlates surface thermography with liver steatosis and inflammation in mice. Future development of this screening tool may improve our ability to study, diagnose and treat liver disease.

Thermal Monitoring of Tumor and Tissue State during Radiation Therapy - A Complex Case of Radiation Recall.

Common radiation dermatitis over radiation fields can be mild as minor erythema but can also be associated with blisters and skin desquamation. This phenomenon has been widely investigated and documented, especially in breast cancer patients. Obesity, smoking, and diabetes are known risk factors; however, we cannot predict the severity of radiation dermatitis prior to treatment. The overwhelming radiation recall dermatitis is an acute inflammatory reaction confined to previously irradiated areas that can be triggered when chemotherapy agents are administered after radiotherapy. This rare, painful skin reaction leads to treatment cessation or alteration. In this study, we investigate the feasibility of using thermography as a tool to predict the response of normal breast tissue and skin to radiation therapy and the risk of developing radiation recall dermatitis. Six women with viable in-breast tumor (breast cancer) and eight women who underwent tumor resection (lumpectomy) were monitored by a thermal camera prior to radiotherapy treatment (breast region) and on weekly basis, in the same environmental conditions, through the radiation course of treatment. One patient developed radiation recall dermatitis when treated with chemotherapy following radiation therapy, and needed intensive local treatments and narcotics with full recovery thereafter. Clinical and treatment data as well as response to radiation were collected prospectively. The ongoing thermal changes observed during the radiation treatment for all patients, with and without viable tumor in the breast, were documented, analyzed, and reported here with detailed comparison to the recognized data for the patient diagnosed with radiation recall dermatitis.

Non-invasive thermal imaging of cardiac remodeling in mice.

Thermal infrared imaging has been suggested as a non-invasive alternative to monitor physiological processes and disease. However, the use of this technique to image internal organs, such as the heart, has not yet been investigated. We sought to determine the ability of our novel thermal image-processing algorithm to detect structural and functional changes in a mouse model of hypertension and cardiac remodeling. Twelve mice were randomly assigned to receive either the pro-inflammatory, hypertensive hormone angiotensin-II (2 mg/kg/day, n = 6) or saline (n = 6) infusion for 28 days. We performed weekly blood pressure measurements, together with serial trans-thoracic echocardiography studies and histopathological evaluation of the hearts. Thermal images were captured with a commercially available thermal camera, and images were processed by our novel algorithm which analyzes relative spatial temperature variation across the animal's thorax. We assessed cardiac inflammation by measuring inflammatory cell infiltration through flow cytometry. Angiotensin infusion increased blood pressure together with cardiac hypertrophy and fibrosis. Thermal imaging at day 28 of the experiment detected an increase in the fraction of the skin heated by the heart in angiotensin-treated mice. Thermal image findings were significantly correlated to left ventricular volume and mass parameters seen on echocardiography (r = 0.8, p < 0.01 and r = 0.6, p = 0.07). We also identified distinct changes in the spatial heat profiles of all angiotensin-treated hearts, possibly reflecting remodeling processes in the hypertensive heart. Finally, a machine learning based model using thermal imaging parameters predicted intervention status in 10 out of 11 mice similar to a model using echocardiographic measurements. Our findings suggest, for the first time, that a new thermal image-processing algorithm successfully correlates surface thermography with cardiac structural changes in mice with hypertensive heart disease.

Automatic Characterization of Retinal Blood Flow Using OCT Angiograms.

PURPOSE: To quantitatively characterize the retinal vascular network in healthy and pathological cases using optical coherence tomography angiography (OCTA) images. METHODS: The study included 56 eyes of 28 patients as follows: 26 healthy, 20 with diabetic retinopathy (DR), 6 with age-related macular degeneration (AMD), and 4 with retinal vein occlusion (RVO). For 33 eyes (16 healthy and 17 with DR), vessel density maps were provided by the OCTA machine. An automatic algorithm classified the image (as healthy, DR, AMD, or RVO) and provided quantitative information obtained from the angiograms, including global vessel density, global fractal dimension, and fovea avascular zone (FAZ) area. Classification results were compared with the diagnosis made by a retina specialist. The quantitative values were compared with the literature and to values provided by the OCTA machine. RESULTS: The success rate of classification was 83.9%. Vessel densities obtained by our algorithm (in healthy and DR cases) were significantly lower than the values reported in previous studies using OCTA. Similarly, they were much lower than the values provided by the OCTA machine. However, vessel densities in the healthy cases were similar to or higher than (depending on the retinal layer) the recently published values that may be considered as gold standard. Our values of fractal dimension were similar to those previously reported. CONCLUSIONS: Our algorithm provides significantly improved vessel density values compared with previous studies. We believe our algorithm successfully omits false vessels. TRANSLATIONAL RELEVANCE: Accurately assessing retinal vessel density enables better evaluation of retinal disorders.

Thermal imaging as a tool for evaluating tumor treatment efficacy.

Breast cancer is the most frequently diagnosed cancer among women in the Western world. Thermography is a nonionizing, noninvasive, portable, and low-cost method that can be used in an outpatient clinic. It was tried as a tool to detect breast cancer tumors, however, it had too many false readings. Thermography has been extensively studied as a breast cancer detection tool but was not used as a treatment monitoring tool. The purpose of this study was to investigate the possibility of using thermal imaging as a feedback system to optimize radiation therapy. Patients were imaged with a thermal camera prior and throughout the radiotherapy sessions. At the end of the session, the images were analyzed for temporal vasculature changes through vessels segmentation image processing tools. Tumors that were not responsive to treatment were observed before the radiation therapy sessions were concluded. Assessing the efficacy of radiotherapy during treatment makes it possible to change the treatment regimen, dose, and radiation field during treatment as well as to individualize treatment schedules to optimize treatment effectiveness.

An automatic approach for morphological analysis and malignancy evaluation of ovarian masses using B-scans.

Ovarian masses are a common phenomenon among women of all ages. The importance of prompt diagnosis of ovarian malignancies is obvious, due to the high mortality rate and the difficulty to detect a tumor in its early stages. In this work, an automatic technique for quantitative analysis and malignancy detection of ovarian masses using B-scan ultrasound (US) images is presented. The core of the technique is morphologic analysis of the ovarian mass. The method employed for this task is divided into two major stages: initial classification of the mass (into one of the three major tumor types: cyst, semisolid, solid), and detailed analysis of the mass. Malignancy evaluation is performed based on the collected data and the criteria provided by commonly used scoring systems. The results reflect adequate performance of the automatic method developed (referring to clinical requirements).