China
Africa
Explore chapters and articles related to this topic
Automated Prostate Image Recognition and Segmentation
Published in Ayman El-Baz, Gyan Pareek, Jasjit S. Suri, Prostate Cancer Imaging, 2018
Ke Yan, Xiuying Wang, Jinman Kim, Changyang Li, Dagan Feng, Mohamed Khadra
In clinical practice, when a patient is deemed at risk of prostate cancer after blood test or rectal examination, the patient is generally required to have a biopsy to confirm diagnosis. A urologist will perform biopsy of the patient to cut out small samples (typically 12 sites) of tissue from the prostate and a pathologist will identify whether the cells are malignant or benign [3]. However, prostate biopsy is an invasive, painful procedure, has a significant risk of infection, and is subject to false-negative results, in particular, when none of the samples are cut from the lesion areas [3]. Alternative noninvasive, biomedical imaging-based methods have recently been widely explored for prostate cancer assessment [4–7]. Imaging-based prostate cancer assessment, computer-aided diagnosis (CAD) systems can be employed to provide automated approaches to the recognition and segmentation of the prostate and this is a fundamental requirement in the analysis of prostate cancer; recognition is to identify the site of the prostate in the image volume and segmentation is used to quantify the volume of the prostate [8]. This approach has the potential to negate the subjectivity and reduce the time needed in manual segmentation and recognition of the prostate [1].
Urology
Published in David A Lisle, Imaging for Students, 2012
An increasing PSA level in an individual patient is considered more significant than a single reading. If either DRE or PSA are abnormal, prostate biopsy is performed. This is done under transrectal ultrasound (TRUS) guidance (Fig. 5.11). Less commonly, biopsies may be performed using a US-guided transperineal approach under general anaesthetic.
The current status of breakthrough devices designation in the United States and innovative medical devices designation in Korea for digital health software
Published in Expert Review of Medical Devices, 2022
Jae Hyun Woo, Eun Cheol Kim, Sung Min Kim
SiMD and SaMD accounted for a total of 26 (8.4%) of 309 designated and published data. Detailed product summaries and major indications for use and approval status are shown in Table 3. In addition, as a result of investigating FDA Approval or Clearance by 31 December 2021, two cases of ‘Paige Prostate’ (Paige. AI, De Novo) and ‘Koios DS’ (Koios Medical, 510(k)) received marketing authorization. Paige Prostate has been classified under the Software Algorithm Device To Assist Users In Digital Pathology (item code QPN) [43]. This was the first AI-based software designed to identify an area of interest on a prostate biopsy image with the highest likelihood of harboring cancer for further review by a pathologist. Authorization for Paige Prostate was based on a clinical study in which 16 pathologists scrutinized 527 prostate biopsy slides. The software improved the pathologists’ ability to detect cancer on individual slide images by an average of 7.3% (from 89.5% to 96.8%) [44]. Koios DS has been classified as Computer-Assisted Diagnostic Software for Lesions Suspicious For Cancer (item code POK) [45]. It is an AI-based software platform used to diagnose thyroid and breast cancer. It increased the thyroid cancer detection rate to 14% and simultaneously reduced false-positive biopsy orders by more than 35%. It decreased interpretation variability by more than 50%. It also decreased the time spent per case by 24% [46].
Use of multiparametric magnetic resonance imaging (mpMRI) in localized prostate cancer
Published in Expert Review of Medical Devices, 2020
Luke O’Connor, Alex Wang, Stephanie M. Walker, Nitin Yerram, Peter A. Pinto, Baris Turkbey
As mentioned earlier, an important question that remains to be answered is whether the presence of a nMRI finding obviates the need for prostate biopsy. However, there is concern that using mpMRI as a triage test to preclude prostate biopsy may lead to a delay in the treatment of clinically significant MRI “invisible“ lesions. For this reason, the true NPV of nMRI remains the cornerstone of this strategy. A meta-analysis of 48 studies performed by Moldovan et al. showed that the NPV greatly varies between studies depending on study design, cancer prevalence, and definitions of mpMRI and clinically significant cancer, thus demonstrating why finding a true NPV for nMRI is difficult in the current literature [31]. Another study performed by Borofsky et al. specifically looked at the number of clinically significant lesions missed on mpMRI [32]. Out of 162 lesions, 26 lesions were missed on mpMRI (one GG 3 tumor, 7 GG 4 tumors, and one GG 5 tumor). Of these missed lesions, 58% were either not seen or characterized as benign based on PI-RADS score (category 1 or 2) [32].
Automatic pathology of prostate cancer in whole mount slides incorporating individual gland classification
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2019
Sabrina Rashid, Guy Nir, Ladan Fazli, Alexander H. Boag, D. Robert Siemens, S. Larry Goldenberg, Purang Abolmaesumi, Septimiu E. Salcudean
Prostate cancer (PCa) is one of the most frequently diagnosed cancers and ranks high among the total cancer related deaths of men worldwide stat. The usual PCa screening process involves a prostate specific antigen test and/or a digital rectal examination. Anomalies in these tests lead clinicians to conduct prostate biopsy. Examination of the microscopic biopsy specimens by pathologists is required for confirming the diagnosis of malignancy and guiding the treatment (Zhu et al. 2006). In case of localised cancers, surgeons often perform radical prostatectomy (RP) on patients, i.e. surgical removal of the entire prostate. The histopathology slices obtained from the cross section of these ex vivo prostates are termed as whole mount (WM) slides. A typical prostate WM slide can be seen in Figure 1. The black contour is the coarse annotation marked by the pathologist on the slide before digitisation.