Expert System for Bone Scan Interpretation Improves Progression Assessment in Bone Metastatic Prostate Cancer

A randomised, double-blind, dose-finding, phase II multicentre study of ODX in the treatment of patients with castration-resistant prostate cancer and skeletal metastases

AIMS

This study aimed to assess the efficacy and safety of ODX, a novel, cytotoxic, bone-targeting drug candidate, in castration-resistant prostate cancer bone metastatic disease.

METHODS

Patients with progressive disease were randomised to ten cycles of ODX, intravenous infusion Q2W (3, 6, and 9 mg/kg, respectively). The primary objective was to assess the relative change from baseline in bone alkaline phosphatase (B-ALP) and serum-aminoterminal-propeptide of Type I procollagen (S-P1NP) at 12 weeks. The inclusion criteria selected were broad, and a double-blind design was used to ensure objective recruitment of patients for the assessment of efficacy. None of the patients received bone-protecting agents during the ODX treatment period.

RESULTS

Fifty-five 21,20 and 14) patients were randomised to ODX (3, 6 and 9 mg/kg), respectively. The lower number of patients in arm 3 was due to too low a recruitment rate towards the end of the study. The median treatment time were 14, 13 and 14 weeks, respectively. The decrease in B-ALP at 12 weeks in study arms 3, 6 and 9 mg/kg was seen in 6/15 (40%), 8/12 (67%) and 5/12 (42%) patients, respectively, whereas the corresponding numbers for P1NP were 8/15 (53%), 8/12 (67%), and 4/12 (33%), respectively. The median decrease in B-ALP and P1NP at 12 weeks for study arms 3, 6 and 9 mg/kg were 37%, 14% and 43%, respectively, and 51%, 40% and 64%, respectively. The decrease in serum C-terminal telopeptide at 12 weeks was seen in the vast majority of patients and in about one-third of patients in bone scan index. ODX was well tolerated, and no drug-related serious adverse events occurred. There were no significant differences between study arms regarding efficacy and safety.

CONCLUSIONS

ODX was well tolerated and demonstrated inhibitory effects on markers related to the vicious cycle in bone at all three doses. The reduction in metastatic burden, assessed with bone scan index, supports this finding. Studies with continued ODX treatment until disease progression are being planned (ClinicalTrials.gov Identifier: NCT02825628).

Application of SPECT and PET / CT with computer-aided diagnosis in bone metastasis of prostate cancer: a review

Bone metastasis has a significant influence on the prognosis of prostate cancer(PCa) patients. In this review, we discussed the current application of PCa bone metastasis diagnosis with single-photon emission computed tomography (SPECT) and positron emission tomography/computed tomography (PET/CT) computer-aided diagnosis(CAD) systems. A literature search identified articles concentrated on PCa bone metastasis and PET/CT or SPECT CAD systems using the PubMed database. We summarized the previous studies focused on CAD systems and manual quantitative markers calculation, and the coincidence rate was acceptable. We also analyzed the quantification methods, advantages, and disadvantages of CAD systems. CAD systems can detect abnormal lesions of PCa patients’ 99mTc-MDP-SPECT, 18F-FDG-PET/CT, 18F-NaF-PET/CT, and 68 Ga-PSMA PET/CT images automated or semi-automated. CAD systems can also calculate the quantitative markers, which can quantify PCa patients’ whole-body bone metastasis tumor burden accurately and quickly and give a standardized and objective result. SPECT and PET/CT CAD systems are potential tools to monitor and quantify bone metastasis lesions of PCa patients simply and accurately, the future clinical application of CAD systems in diagnosing PCa bone metastasis lesions is necessary and feasible.

Intelligent Fault Diagnosis Method for Blade Damage of Quad-Rotor UAV Based on Stacked Pruning Sparse Denoising Autoencoder and Convolutional Neural Network

This paper introduces a new intelligent fault diagnosis method based on stack pruning sparse denoising autoencoder and convolutional neural network (sPSDAE-CNN). This method processes the original input data by using a stack denoising autoencoder. Different from the traditional autoencoder, stack pruning sparse denoising autoencoder includes a fully connected autoencoding network, the features extracted from the front layer of the network are used for the operation of the subsequent layer, which means that some new connections will appear between the front and rear layers of the network, reduce the loss of information, and obtain more effective features. Firstly, a one-dimensional sliding window is introduced for data enhancement. In addition, transforming one-dimensional time-domain data into the two-dimensional gray image can further improve the deep learning (DL) ability of models. At the same time, pruning operation is introduced to improve the training efficiency and accuracy of the network. The convolutional neural network model with sPSDAE has a faster training speed, strong adaptability to noise interference signals, and can also suppress the over-fitting problem of the convolutional neural network to a certain extent. Actual experiments show that for the fault of unmanned aerial vehicle (UAV) blade damage, the sPSDAE-CNN model we use has better stability and reliable prediction accuracy than traditional convolutional neural networks. At the same time, For noise signals, better results can be obtained. The experimental results show that the sPSDAE-CNN model still has a good diagnostic accuracy rate in a high-noise environment. In the case of a signal-to-noise ratio of −4, it still has an accuracy rate of 90%.

Prognosis of patients with prostate cancer and bone metastasis from the Japanese Prostatic Cancer Registry of Standard Hormonal and Chemotherapy Using Bone Scan Index cohort study

Objective

To determine prognostic factors including the Bone Scan Index in prostate cancer patients receiving standard hormonal therapy and chemotherapy.

Methods

This multicenter Prostatic Cancer Registry of Standard Hormonal and Chemotherapy Using Bone Scan Index study involved 30 hospitals and enrolled 247 patients (age 71 ± 8 years) with metastatic hormone‐sensitive prostate cancer (n = 148) under hormone therapy and metastatic castration‐resistant prostate cancer (n = 99) under chemotherapy. The Bone Scan Index (%) was determined by whole‐body bone scintigraphy using ^99mTc‐methylenediphosphonate. Patients were classified into tertiles and binary groups, and predictors of all‐cause death including Bone Scan Index, prostate‐specific antigen, and bone metabolic markers were determined using survival and proportional hazard analyses.

Results

During a mean follow‐up period of 716 ± 404 days, 81 (33%) of the patients died, and 3‐year mortality rates were 20% and 52% in the metastatic hormone‐sensitive prostate cancer and metastatic castration‐resistant prostate cancer groups, respectively. Survival analysis showed that a Bone Scan Index >3.5% was a significant determinant of death in the metastatic hormone‐sensitive prostate cancer group, whereas prostate‐specific antigen >55 ng/mL before chemotherapy was a determinant of prognosis in the metastatic castration‐resistant prostate cancer group. A Bone Scan Index >3.5% was also associated with a high incidence of prostate‐specific antigen progression in the metastatic hormone‐sensitive prostate cancer group. Patients with metastatic hormone‐sensitive prostate cancer and a better Bone Scan Index response (>45%) to treatment had lower mortality rates than those without such response.

Conclusion

The Bone Scan Index and hot spot number are significant determinants of 3‐year mortality, and combining the Bone Scan Index with prostate‐specific antigen should contribute to the management of prostate cancer patients with bone metastasis.

Artificial Intelligence in Clinical Decision Support: a Focused Literature Survey

OBJECTIVES

This survey analyses the latest literature contributions to clinical decision support systems (DSSs) on a two-year period (2017-2018), focusing on the approaches that adopt Artificial Intelligence (AI) techniques in a broad sense. The goal is to analyse the distribution of data-driven AI approaches with respect to "classical" knowledge-based ones, and to consider the issues raised and their possible solutions.

METHODS

We included PubMed and Web of ScienceTM publications, focusing on contributions describing clinical DSSs that adopted one or more AI methodologies.

RESULTS

We selected 75 papers, 49 of which describe approaches in the data-driven AI area, 20 present purely knowledge-based DSSs, and 6 adopt hybrid approaches relying on both formalized knowledge and data.

CONCLUSIONS

Recent studies in the clinical DSS area demonstrate a prevalence of data-driven AI, which can be adopted autonomously in purely data-driven systems, or in cooperation with domain knowledge in hybrid systems. Such hybrid approaches, able to conjugate all available knowledge sources through proper knowledge integration steps, represent an interesting example of synergy between the two AI categories. This synergy can lead to the resolution of some existing issues, such as the need for transparency and explainability, nowadays recognized as central themes to be addressed by both AI and medical informatics research.

Measuring the unmeasurable: automated bone scan index as a quantitative endpoint in prostate cancer clinical trials

BACKGROUND

Up to 90% of men with metastatic castration-resistant prostate cancer (mCRPC) will have a distribution of disease that includes bone metastases demonstrated on a Technetium-99m (^99mTc-MDP) bone scan. The Prostate Cancer Working Group 2 and 3 Consensus Criteria standardized the criteria for assessing progression based on the development of new lesions. These criteria have been recognized by regulatory authorities for drug approval. The bone scan index (BSI) is a method to quantitatively measure the burden of bony disease, and can assess both disease progression and regression. The automated BSI (aBSI) is a method of computer analysis to assess BSI, and is being qualified as a clinical trials endpoint.

METHODS

Manual searching was used to identify the literature on BSI and aBSI. We summarize the most relevant aspects of the retrospective and prospective studies evaluating aBSI measurements, and provide a critical discussion on the potential advantages and caveats of aBSI.

RESULTS

The development of neural artificial networks (EXINI boneBSI) to automatically determine the BSI reduces the turnaround time for assessing BSI with high reproducibility and accuracy. Several studies showed that the concordance between aBSI and BSI, as well as the interobserver concordance of aBSI, was >0.95. In a phase 3 assessment of aBSI, a doubling value increased the risk of death in 20%, pre-treatment aBSI values independently correlated with overall survival (OS) and time to symptomatic progression. Retrospective studies suggest that a decrease in aBSI after treatment may correlate with higher survival when compared with increasing aBSI.

CONCLUSIONS

aBSI provides a quantitative measurement that is feasible, reproducible, and in analyses to date correlates with OS and symptomatic progression. These findings support the aBSI to risk-stratify men with mCRPC for clinical trial enrollment. Future studies quantifying aBSI change over time as an intermediate endpoint for evaluating new systemic therapies are needed.

EBONI: A Tool for Automated Quantification of Bone Metastasis Load in PSMA PET/CT

Prostate-specific membrane antigen (PSMA) PET/CT has a high diagnostic accuracy for lesion detection in metastatic prostate cancer, including bone metastases. Novel therapeutic approaches require valid biomarkers for standardized disease staging and for evaluation of progression and therapy response. Here, we introduce EBONI (Evaluation of Bone Involvement), a software tool to automatically quantify the bone metastasis load in PSMA PET/CT. Lesion quantity, mean and maximum lesional SUV, z score, and percentage of affected bone volume are determined. EBONI is open source and freely available. Methods: To validate EBONI, the results of automated quantification of 38 PSMA PET/CT scans with different levels of bone involvement were compared with visual expert reading. The influence of SUV threshold and Hounsfield unit thresholds was analyzed. Results: A high correlation between bone lesion quantity as determined visually and automatically was found (SUV_max, r² = 0.97; SUV_mean, r² = 0.88; lesion count, r² = 0.97). The Hounsfield unit threshold had no significant influence, whereas an SUV threshold of 2.5 proved optimal for automated lesion quantification. The systematic error of false-positive tissue misclassification was low, occurred mainly around the salivary and lacrimal glands, and could easily be corrected. There were no false-negative ratings. Conclusion: EBONI analysis is robust, quick (<3 min per scan), and 100% reproducible. It allows rater-independent quantification of bone metastasis in metastatic prostate cancer. It provides lesion quantification equivalent to that of visual assessment, as well as providing complementary information. It can be easily implemented as an add-on to visual analysis of PSMA PET/CT scans and has the potential to reduce turnaround time.