This study aimed to explore efficient ways to diagnose bone metastasis early using bone scintigraphy images through negative mining, pre-training, the convolutional neural network, and deep learning. We studied 205 prostate cancer patients and 371 breast cancer patients and used bone scintigraphy data from breast cancer patients to pre-train a YOLO v4 with a false-positive reduction strategy. With the pre-trained model, transferred learning was applied to prostate cancer patients to build a model to detect and identify metastasis locations using bone scintigraphy. Ten-fold cross validation was conducted. The mean sensitivity and precision rates for bone metastasis location detection and classification (lesion-based) in the chests of prostate patients were 0.72 ± 0.04 and 0.90 ± 0.04, respectively. The mean sensitivity and specificity rates for bone metastasis classification (patient-based) in the chests of prostate patients were 0.94 ± 0.09 and 0.92 ± 0.09, respectively. The developed system has the potential to provide pre-diagnostic reports to aid in physicians’ final decisions.

The objective of this study was to correlate the association between mutated KRAS and wild-type colorectal cancer (CRC) by using various F-FDG PET-related parameters. One hundred twenty-one CRC patients who had undergone preoperative PET/CT were included in this study. Several PET/CT-related parameters, including SUVmax and various thresholds of metabolic tumor volume, total lesion glycolysis, and PET/CT-based tumor width, were measured. Tumor- and PET/CT-related parameters were correlated with genomic expression between KRAS mutant and wild-type groups, using a Mann-Whitney U test and logistic regression analysis. Colorectal cancer tumors with a mutated KRAS exhibited higher SUVmax and an increased accumulation of FDG among several threshold methods. Multivariate analysis showed that SUVmax and using a 40% threshold level for maximal uptake of TW (TW40%) were the 2 predictors of KRAS mutations. The odds ratio was 1.23 for SUVmax (P = 0.02; 95% confidence interval, 1.01-1.52) and 1.15 for TW40% (P = 0.02; 95% confidence interval, 1.02-1.30). The accuracy of SUVmax for predicting mutated KRAS was higher in patients with colon or sigmoid colon cancers, whereas it was TW40% in those with rectal cancers. SUVmax and TW40% were associated in CRC with KRAS mutations. PET/CT parameters can supplement genomic analysis to determine KRAS expression in CRC.