Assessing Trustworthy AI in Times of COVID-19: Deep Learning for Predicting a Multiregional Score Conveying the Degree of Lung Compromise in COVID-19 Patients.

This article's main contributions are twofold: 1) to demonstrate how to apply the general European Union's High-Level Expert Group's (EU HLEG) guidelines for trustworthy AI in practice for the domain of healthcare and 2) to investigate the research question of what does "trustworthy AI" mean at the time of the COVID-19 pandemic. To this end, we present the results of a post-hoc self-assessment to evaluate the trustworthiness of an AI system for predicting a multiregional score conveying the degree of lung compromise in COVID-19 patients, developed and verified by an interdisciplinary team with members from academia, public hospitals, and industry in time of pandemic. The AI system aims to help radiologists to estimate and communicate the severity of damage in a patient's lung from Chest X-rays. It has been experimentally deployed in the radiology department of the ASST Spedali Civili clinic in Brescia, Italy, since December 2020 during pandemic time. The methodology we have applied for our post-hoc assessment, called Z-Inspection®, uses sociotechnical scenarios to identify ethical, technical, and domain-specific issues in the use of the AI system in the context of the pandemic.

Should We Rely on AI to Help Avoid Bias in Patient Selection for Major Surgery?

Many regard iatrogenic injuries as consequences of diagnosis or intervention actions. But inaction-not offering indicated major surgery-can also result in iatrogenic injury. This article explores some surgeons' overestimations of operative risk based on patients' race and socioeconomic status as unduly influential in their decisions about whether to perform major cancer or cardiac surgery on some patients with appropriate clinical indications. This article also considers artificial intelligence and machine learning-based clinical decision support systems that might offer more accurate, individualized risk assessment that could make patient selection processes more equitable, thereby mitigating racial and ethnic inequity in cancer and cardiac disease.

How Should Surgeons Communicate About Palliative and Curative Intentions, Purposes, and Outcomes?

How surgeons describe procedures should be accurate, precise, and concordant with patients' values. By focusing on intention rather than realistic goals, terms like curative and palliative, when applied to high-stakes operations, such as a Whipple pancreaticoduodenectomy, can be confusing to patients. This case commentary argues that surgeons' language choices can influence patients' decisions and experiences.

Who, If Not the FDA, Should Regulate Implantable Brain-Computer Interface Devices?

Implantable brain-computer interface (BCI) and other devices with potential for both therapeutic purposes and human enhancement are being rapidly developed. The distinction between therapeutic and enhancement uses of these devices is not well defined. While the US Food and Drug Administration (FDA) rightly determines what is safe and effective, this article argues that the FDA should not make subjective, value-laden assessments about risks and benefits when it comes to approval of BCIs for therapy and enhancement. This article also argues that determining BCIs' benefits to society requires deliberations on values that the FDA is neither accustomed to making nor qualified to make. Given the inadequacy of the FDA's safe-and-effective standard to judge devices spanning the spectrum of therapy to enhancement, this article argues that BCI regulation should not be overseen by the FDA.

Ethical Rationing of Personal Protective Equipment to Minimize Moral Residue During the COVID-19 Pandemic.

This article proposes systems for the fair distribution of scarce resources to healthcare providers. It builds on classic ethical structures and adapts them to the equitable distribution of personal protective equipment (PPE) to clinicians at risk of contracting novel corona virus-19 (COVID-19). The article also defines systems of allocation that are generally considered unethical and are to be avoided. We emphasize that policies must be transparent, collaborative, applied equally, and have a system of accountability. It is recognized that unless the supply of PPE is quickly replenished, or viable alternatives to traditional equipment are devised in the coming days to weeks, hospitals and healthcare systems will face the difficult task of rationing PPE to at-risk clinicians. This paper suggests an ethical framework for that process.

The molecular basis of pancreatic fibrosis: common stromal gene expression in chronic pancreatitis and pancreatic adenocarcinoma.

OBJECTIVES: Tissue desmoplasia occurs in a number of disease states, but its molecular basis is poorly understood. To determine which genes are overexpressed in cells contained within the desmoplastic stroma of pancreatic adenocarcinoma and chronic pancreatitis, we undertook genetic profiling of microdissected tissue samples of pancreatic adenocarcinoma, chronic pancreatitis, normal pancreas, and pancreatic cancer cell lines. We observed that samples of both pancreatic adenocarcinoma and chronic pancreatitis showed elevated expression of many shared genes compared with the normal pancreas. We hypothesized that these common genes likely important in stromal production and/or function could be identified using a strategy that involved comparisons between pancreatic adenocarcinoma, chronic pancreatitis, normal pancreas, and pancreatic cancer cell lines. METHODS: We performed oligonucleotide microarray analysis of 6800 different genes expressed in 10 samples of pancreatic adenocarcinoma, 5 samples of normal pancreas, 5 samples of chronic pancreatitis, and 7 pancreatic cancer cell lines. Microarray findings were validated with RT-PCR, and immunohistochemistry was used to verify protein localization to the stromal compartment of both pancreatic cancer and chronic pancreatitis. RESULTS: We employed a deductive comparison whereby genes expressed in the normal pancreas and pancreatic cancer cell lines were selectively eliminated from those expressed in common by pancreatic adenocarcinoma and chronic pancreatitis. This strategy identified 107 genes predicted to be expressed within cells of the stromal compartment of both pancreatic adenocarcinoma and chronic pancreatitis. CONCLUSIONS: These genes are likely important factors in epithelial-stromal signaling in pancreatic desmoplasia and may serve as diagnostic or therapeutic targets.

Raf kinase inhibitory protein inhibits beta-cell proliferation.

BACKGROUND: Raf-1 kinase inhibitory protein (RKIP) was recently identified as a physiologic endogenous inhibitor of the extracellular signal-regulated kinase (ERK) pathway. The expression and role of RKIP within the pancreas are unknown. METHODS: RKIP expression in normal pancreas and human insulinomas was examined by using paraffin-embedded sections. Co-localization of RKIP within islet cell subtypes was performed by using double immunofluorescence staining with antibodies directed toward RKIP and endocrine markers. To examine the role of RKIP in beta-cell proliferation, stable expression of sense (ss) and antisense (as) RKIP was established in HIT-T15 beta cells. The effect of RKIP on the ERK-signaling pathway in beta cells was determined by Western blotting with the use of phospho-specific antibodies directed against mitogen-activated protein kinase kinase (MEK) and ERK. The role of RKIP in beta-cell proliferation was assessed by using MTS assay and FACS analysis. RESULTS: RKIP was expressed only within pancreatic islet cells. Immunofluorescent double staining revealed that RKIP was expressed in most beta cells and a subset of pancreatic polypeptide-expressing cells. Based on the known function of RKIP, we hypothesized that RKIP expression would be downregulated in insulinomas: 8 of 9 human insulinomas demonstrated no RKIP staining, with decreased expression in 1 of 9 insulinomas. Studies using asRKIP and ssRKIP demonstrated that RKIP blocked activation of MEK and ERK by Raf-1 in beta cells. We also showed that RKIP inhibited beta-cell proliferation by altering cell cycle distribution, rather than by promoting apoptosis. CONCLUSIONS: RKIP is important in beta-cell proliferation, and its downregulation may play a role in islet neoplasia.

A transforming growth factor beta-induced Smad3/Smad4 complex directly activates protein kinase A.

Transforming growth factor beta (TGFbeta) interacts with cell surface receptors to initiate a signaling cascade critical in regulating growth, differentiation, and development of many cell types. TGFbeta signaling involves activation of Smad proteins which directly regulate target gene expression. Here we show that Smad proteins also regulate gene expression by using a previously unrecognized pathway involving direct interaction with protein kinase A (PKA). PKA has numerous effects on growth, differentiation, and apoptosis, and activation of PKA is generally initiated by increased cellular cyclic AMP (cAMP). However, we found that TGFbeta activates PKA independent of increased cAMP, and our observations support the conclusion that there is formation of a complex between Smad proteins and the regulatory subunit of PKA, with release of the catalytic subunit from the PKA holoenzyme. We also found that the activation of PKA was required for TGFbeta activation of CREB, induction of p21(Cip1), and inhibition of cell growth. Taken together, these data indicate an important and previously unrecognized interaction between the TGFbeta and PKA signaling pathways.

Molecular profiling of pancreatic adenocarcinoma and chronic pancreatitis identifies multiple genes differentially regulated in pancreatic cancer.

The molecular basis of pancreatic cancer is not understood. Previous attempts to determine the specific genes expressed in pancreatic cancer have been hampered by similarities between adenocarcinoma and chronic pancreatitis. In the current study, microarrays (Affymetrix) were used to profile gene expression in pancreatic adenocarcinoma (10), pancreatic cancer cell lines (7), chronic pancreatitis (5), and normal pancreas (5). Molecular profiling indicated a large number of genes differentially expressed between pancreatic cancer and normal pancreas but many fewer differences between pancreatic cancer and chronic pancreatitis, likely because of the shared stromal influences in the two diseases. To specifically identify genes expressed in neoplastic epithelium, we selected genes more highly expressed (>2-fold, p < 0.01) in adenocarcinoma compared with both normal pancreas and chronic pancreatitis and which were also highly expressed in pancreatic cancer cell lines. This strategy yielded 158 genes, of which 124 were not previously associated with pancreatic cancer. Quantitative-reverse transcription-PCR for two molecules, S100P and 14-3-3sigma, validated the microarray data. Support for the success of the neoplastic cell gene expression identification strategy was obtained by immunocytochemical localization of four representative genes, 14-3-3sigma, S100P, S100A6, and beta4 integrin, to neoplastic cells in pancreatic tumors. Thus, comparisons between pancreatic adenocarcinoma, pancreatic cancer cell lines, normal pancreas, and chronic pancreatitis have identified genes that are selectively expressed in the neoplastic epithelium of pancreatic adenocarcinoma. These data provide new insights into the molecular pathology of pancreatic cancer that may be useful for detection, diagnosis, and treatment.

Unusual causes of benign biliary strictures with cholangiographic features of cholangiocarcinoma.

Focal strictures occurring at the hepatic duct confluence, or within the common hepatic duct or common bile duct in patients without a history of prior surgery in that region or stone disease, are usually thought to represent cholangiocarcinoma until proved otherwise. However, not uncommonly, patients undergo surgical exploration for a preoperative diagnosis of cholangiocarcinoma, based on the cholangiographic appearance of the lesion, only to find histologically that the stricture was benign in nature. Despite sophisticated radiographic, endoscopic, and histologic studies, it is often impossible before laparotomy to distinguish malignant from benign strictures when they have the characteristic radiographic appearance of cholangiocarcinoma. Even at the risk of overtreating some benign cases, most agree that aggressive surgical resection is the treatment of choice, given the serious consequences resulting from a failure to diagnose and adequately treat cholangiocarcinoma. Four patients who presented to our institution between February 1991 and June 2000 underwent laparotomy for a preoperative diagnosis of biliary tract malignancy based on clinical presentation and cholangiographic findings. The final pathology report in all patients showed marked fibrosis and inflammation of the biliary duct without evidence of malignancy. A review of the patient data and the relevant literature identified benign causes of focal extrahepatic biliary strictures associated with concomitant disease processes in two of the four patients. We present these cases and discuss the benign etiologies with emphasis on the role of surgery in both diagnosis and treatment.