Deep-Learning-Based Smartphone Application for Self-Diagnosis of Scleral Jaundice in Patients with Hepatobiliary and Pancreatic Diseases

Abstract

Outpatient detection of total bilirubin levels should be performed regularly to monitor the recurrence of jaundice in hepatobiliary and pancreatic disease patients. However, frequent hospital visits for blood testing are burdensome for patients with poor medical conditions. This study validates a novel deep-learning-based smartphone application for the self-diagnosis of scleral jaundice in such patients. The system predicts total serum bilirubin levels using the deep-learning-based regression analysis of scleral photos taken by the smartphone’s built-in camera. Enrolled patients were randomly assigned to either the training cohort (n = 90, 1034 photos) or the validation cohort (n = 40, 426 photos). The intraclass correlation coefficient value for predicted serum total bilirubin (PSB) derived from the images repeatedly taken at the same time for the same patient showed good reliability (0.86). A strong correlation between measured serum total bilirubin (MSB) and PSB was observed in the subgroup with MSB levels ≥1.5 mg/dL (Spearman rho = 0.70, p < 0.001). The receiver operating characteristic curve for PSB showed that the area under the curve was 0.93, demonstrating good test performance as a predictor of hyperbilirubinemia (p < 0.001). Using a cut-off PSB ≥1.5, the prediction sensitivity of hyperbilirubinemia was 80.0%, with a specificity of 92.6%. Hence, the tool is effective for patient monitoring.

Identification and Quantification of Jaundice by Trans-Conjunctiva Optical Imaging Using a Human Brain-like Algorithm: A Cross-Sectional Study

“…Park et al recently investigated a deep-learning system for detecting jaundice caused by hepatobiliary and pancreatic diseases using a smartphone [11]. This report also demonstrated excellent hyperbilirubinemia (TSB ≥ 1.5 mg/dL) prediction sensitivity (80.0%) and specificity (92.6%).…”

“…Most other studies using smartphones maintained consistent conditions by fixing the light source and distance or employing a color consistency patch [7][8][9][10][11]19,[21][22][23][24]. Typically, cameras modify color information via their built-in auto white balance function to minimize the effect of ambient light colors, ensuring that captured images appear natural when viewed by human observers.…”

“Notification! You May Have Cancer.” Could Smartphones and Wearables Help Detect Cancer Early?

“…Changes in voice such as hoarseness (associated with head and neck cancer and lung cancer) [14] GPS location tracking and activity tracking [15] • Reduced activity resulting from fatigue (associated with multiple cancers) [14] Image capture and analysis [16][17][18] • Anemia detected from images of the skin or eyes (associated with multiple cancers) [19] • Jaundice detected from images of the skin or eyes (associated with pancreatic cancer) [20] • Changes in skin lesions (associated with skin cancer) [14] Temperature measurement [21] • Rise in temperature (associated with pancreatic cancer) [22] Body composition using image analysis and electro dermal activity [23] • Weight loss (associated with multiple cancers) [14] Photoplethysmogram [24] • Anemia (associated with multiple cancers) [19] Sensors could allow the detection of changes prior to them being noticed or interpreted as symptoms, for example, a reduction in activity prior to fatigue or changes in food consumption prior to weight loss. There is recent evidence that monitoring day-to-day purchases could detect an increase in over-the-counter pain and indigestion medication 8 months prior to ovarian cancer diagnosis [25].…”

Portable Smartphone-Enabled Single-Drop Sensors Based on CsPbX3 (X = Cl, Br) Perovskite Composite Nanofiber Membranes

“…Most smartphones already possess resolutions of over 24 megapixels for taking photographs. − This high-resolution photographs can also be transmitted to a server computer via a 5G network in quick time followed by digital image processing and analysis. − This makes it possible to detect the RGB color ratio based on photographs from portable smartphones. , On the other hand, test strips are usually used to detect low concentrations of liquid, but most of them cannot be reused. , For example, pH test paper has to be contaminated by the target solution. − However, this direct contact may lead to contamination of the target solution, which leads to this target solution not being able to be reused either . Therefore, using a reusable test strip with a smartphone and only one drop of liquid to achieve detection has great advantages.…”