Automatic Liver Viability Scoring with Deep Learning and Hyperspectral Imaging

Hyperspectral imaging (HSI) is a non-invasive imaging modality already applied to evaluate hepatic oxygenation and to discriminate different models of hepatic ischemia. Nevertheless, the ability of HSI to detect and predict the reperfusion damage intraoperatively was not yet assessed. Hypoxia caused by hepatic artery occlusion (HAO) in the liver brings about dreadful vascular complications known as ischemia-reperfusion injury (IRI). Here, we show the evaluation of liver viability in an HAO model with an artificial intelligence-based analysis of HSI. We have combined the potential of HSI to extract quantitative optical tissue properties with a deep learning-based model using convolutional neural networks. The artificial intelligence (AI) score of liver viability showed a significant correlation with capillary lactate from the liver surface (r = −0.78, p = 0.0320) and Suzuki’s score (r = −0.96, p = 0.0012). CD31 immunostaining confirmed the microvascular damage accordingly with the AI score. Our results ultimately show the potential of an HSI-AI-based analysis to predict liver viability, thereby prompting for intraoperative tool development to explore its application in a clinical setting.

Multi-organ system for the evaluation of efficacy and off-target toxicity of anticancer therapeutics

A pumpless, reconfigurable, multi-organ–on–a–chip system containing recirculating serum-free medium can be used to predict preclinical on-target efficacy, metabolic conversion, and measurement of off-target toxicity of drugs using functional biological microelectromechanical systems. In the first configuration of the system, primary human hepatocytes were cultured with two cancer-derived human bone marrow cell lines for antileukemia drug analysis in which diclofenac and imatinib demonstrated a cytostatic effect on bone marrow cancer proliferation. Liver viability was not affected by imatinib; however, diclofenac reduced liver viability by 30%. The second configuration housed a multidrug-resistant vulva cancer line, a non–multidrug-resistant breast cancer line, primary hepatocytes, and induced pluripotent stem cell–derived cardiomyocytes. Tamoxifen reduced viability of the breast cancer cells only after metabolite generation but did not affect the vulva cancer cells except when coadministered with verapamil, a permeability glycoprotein inhibitor. Both tamoxifen alone and coadministration with verapamil produced off-target cardiac effects as indicated by a reduction of contractile force, beat frequency, and conduction velocity but did not affect viability. These systems demonstrate the utility of a human cell–based in vitro culture system to evaluate both on-target efficacy and off-target toxicity for parent drugs and their metabolites; these systems can augment and reduce the use of animals and increase the efficiency of drug evaluations in preclinical studies.