Checklist for Reproducibility of Deep Learning in Medical Imaging.

The application of deep learning (DL) in medicine introduces transformative tools with the potential to enhance prognosis, diagnosis, and treatment planning. However, ensuring transparent documentation is essential for researchers to enhance reproducibility and refine techniques. Our study addresses the unique challenges presented by DL in medical imaging by developing a comprehensive checklist using the Delphi method to enhance reproducibility and reliability in this dynamic field. We compiled a preliminary checklist based on a comprehensive review of existing checklists and relevant literature. A panel of 11 experts in medical imaging and DL assessed these items using Likert scales, with two survey rounds to refine responses and gauge consensus. We also employed the content validity ratio with a cutoff of 0.59 to determine item face and content validity. Round 1 included a 27-item questionnaire, with 12 items demonstrating high consensus for face and content validity that were then left out of round 2. Round 2 involved refining the checklist, resulting in an additional 17 items. In the last round, 3 items were deemed non-essential or infeasible, while 2 newly suggested items received unanimous agreement for inclusion, resulting in a final 26-item DL model reporting checklist derived from the Delphi process. The 26-item checklist facilitates the reproducible reporting of DL tools and enables scientists to replicate the study's results.

[The effect of portal blood stasis on lung and renal injury induced by hepatic ischemia reperfusion in a rabbit model].

OBJECTIVE: To investigate the effect and mechanism of portal blood stasis on lung and renal injury induced by hepatic ischemia reperfusion. METHODS: A rabbit hepatic ischemia reperfusion injury model was established by hepatic portal occlusion and in situ hypothermic irrigation for 30 min. Twenty-four New Zealand white rabbits were employed and randomly divided into 3 groups equally by different dosage of portal blood stasis removal: group A5 (5 ml blood removal), group A10 (10 ml blood removal),and group B (no blood removal). Eight rabbits were served as controls with no hepatic portal occlusion and hypothermic irrigation. After reperfusion 4 h serum endotoxin content, tumor necrosis factor-alpha (TNF-alpha), urea nitrogen (BUN), and creatinine (Cr) were examined respectively, meantime lung and kidney tissues were sampled to determine the content of malondialdehyde (MDA), superoxide dismutase (SOD), the pathology, and wet to dry weight ratio, broncho-alveolar lavage fluid protein content in lung tissues. RESULTS: Removing portal blood stasis ameliorated lung and renal injury as shown by decreasing the level of serum endotoxin, TNF-alpha, BUN, Cr, wet to dry weight ratio, broncho-alveolar lavage fluid protein content, MDA, SOD. TNF-alpha, Cr, broncho-alveolar lavage fluid protein content in lung tissues and MDA in kidney tissue in group A5 were significantly reduced compared with those in group B (P < 0.05), while in lung tissue in group A10 were also markedly reduced (P < 0.05). The activation of SOD in group A5 were significantly increased (P < 0.05). CONCLUSIONS: Removal of portal blood stasis before the resume of splanchnic circulation may ameliorate the lung and renal injury induced by hepatic ischemia reperfusion. The possible mechanism may be that portal blood stasis removal reduces endotoxin absorption, and further decreases production of serum TNF-alpha.

[Effect of the amount of portal blood stasis removal on endotoxemia and liver function after liver transplantation].

OBJECTIVE: To investigate the influence of the amount of portal blood stasis removal on endotoxemia and liver function after liver transplantation. METHODS: Forty-seven patients who received liver transplantation from February 2006 to November 2007 were divided into 2 groups according to the amount of portal blood stasis removal during operation: group A (n = 26) 50 ml and group B (n = 21) 200 ml of portal blood stasis removal respectively. The levels of plasma endotoxin, D-lactate, tumor necrosis factor-alpha, interleukin-6, liver function and blood coagulation were examined and analyzed. RESULTS: Under the condition of no significant difference in sex, age, primary liver diseases and Child-pugh's classification, cold ischemic time, total operation and anhepatic time, operation methods, volume of blood loss and transfusion, and all preoperative observations. Most of observations showed the restoration of the patients in group B was better than that in group A. The plasma levels of endotoxin, D-lactate, tumor necrosis factor-alpha, interleukin-6, alanine aminotransferase, aspartate aminotransferase, prothrombin time and activated partial thromboplastin time in group B were significantly lower than those in group A (P < 0.05). The level of plasma prealbumin in group B was significantly higher than that in group A (P < 0.05). CONCLUSIONS: The removal of 200 ml portal blood stasis leads to a better results than that of 50 ml, and it can help alleviate endotoxemia and facilitate the restoration of the liver function after liver transplantation.

[A study of interferon-gamma induced airway mucous cell apoptosis and its mechanisms].

OBJECTIVE: To investigate the effect of interferon-gamma (IFN-gamma) on airway mucous cells and its mechanisms. METHODS: (1) Normal human bronchial epithelial cells (NHBEs) were cultured under specific conditions, and treated by IFN-gamma for 3 days. The cells were analyzed with fluorescein isothiocyanate (FITC) and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay. (2) Twenty wild type C57BL/6J mice were immunized and divided into 4 groups, and treated with IFN-gamma (50 ng and 100 ng, respectively), interleukin-13 (IL-13, 5 microg) and saline by nostril instillation. The mice were sacrificed and the airway mucous cells were analyzed by morphometry and TUNEL assay. RESULTS: (1) IFN-gamma induced apoptosis in NHBEs, which showed condensed nuclei, nuclear and DNA fragmentaion, and were positive by TUNEL assay. Bax was upregulated and translocated from cell plasma to mitochondria under the treatment. (2) Airway mucous cell account in 100 ng IFN-gamma instillation immunized mice group (28 +/- 6 mucous cells/mm basal lamina) was significantly decreased as compared to that in saline (58 +/- 12) and IL-13 (59 +/- 6) instillation groups (all < 0.05). There was no difference among the IFN-gamma 50 ng (48 +/- 11), saline (58 +/- 12) and IL-13 (59 +/- 6) instillation groups (all P > 0.05). TUNEL assay was also positive in airway mucous cells from IFN-gamma instillation mice as compared to saline instillation mice. CONCLUSIONS: IFN-gamma leads to airway mucous cell apoptosis by Bax upregulating and translocation into mitochondria. This might be of significance in the new therapies of asthma.

[Expression of tumor metastasis-suppressor gene KiSS-1 and matrix metalloproteinase-9 in portal vein tumor thrombus of hepatocellular carcinoma].

BACKGROUND & OBJECTIVE: Tumor metastasis-suppressor gene KiSS-1 is related to the metastasis of malignancies. However, its correlation to the metastasis, especially the formation of portal vein tumor thrombus (PVTT), of hepatocellular carcinoma (HCC) has seldom been reported. This study was to investigate the function of KiSS-1 gene in the formation of PVTT of HCC, and explore the machanism. METHODS: The expression of KiSS-1 and matrix metalloproteinase-9 (MMP-9) in 50 specimens of HCC (31 cases with PVTT and 19 cases without PVTT) and 11 specimens of PVTT was detected by reverse transcription-polymerase chain reaction (RT-PCR) and immunohistochemistry (IHC). RESULTS: The positive rates of KiSS-1 mRNA and protein were significantly lower in PVTT group and HCC with PVTT group than in HCC without PVTT group (18.2% and 16.1 % vs. 63.2%, 0% and 12.9% vs. 47.4%, P < 0.05)û the positive rates of MMP-9 mRNA and protein were significantly higher in PVTT group and HCC with PVTT group than in HCC without PVTT group (72.7% and 77.4% vs. 31.6%, 81.8% and 83.9% vs. 42.1%, P < 0.05). There was no significant difference between PVTT group and HCC with PVTT group (P > 0.05). KiSS-1 expression was negatively related to MMP-9 expression (r=-0.362 for mRNA, and r=-0.473 for protein, each P < 0.05). CONCLUSIONS: KiSS-1 and MMP-9 are closely related to the formation of PVTT. KiSS-1 gene might suppress the formation of PVTT by suppressing MMP-9 synthesis.