• Korean Journal of Radiology
• /
• v.23 no.5
• /
• pp.505-516
• /
• 2022

Objective: To evaluate whether artificial intelligence (AI) for detecting breast cancer on mammography can improve the performance and time efficiency of radiologists reading mammograms. Materials and Methods: A commercial deep learning-based software for mammography was validated using external data collected from 200 patients, 100 each with and without breast cancer (40 with benign lesions and 60 without lesions) from one hospital. Ten readers, including five breast specialist radiologists (BSRs) and five general radiologists (GRs), assessed all mammography images using a seven-point scale to rate the likelihood of malignancy in two sessions, with and without the aid of the AI-based software, and the reading time was automatically recorded using a web-based reporting system. Two reading sessions were conducted with a two-month washout period in between. Differences in the area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, and reading time between reading with and without AI were analyzed, accounting for data clustering by readers when indicated. Results: The AUROC of the AI alone, BSR (average across five readers), and GR (average across five readers) groups was 0.915 (95% confidence interval, 0.876-0.954), 0.813 (0.756-0.870), and 0.684 (0.616-0.752), respectively. With AI assistance, the AUROC significantly increased to 0.884 (0.840-0.928) and 0.833 (0.779-0.887) in the BSR and GR groups, respectively (p = 0.007 and p < 0.001, respectively). Sensitivity was improved by AI assistance in both groups (74.6% vs. 88.6% in BSR, p < 0.001; 52.1% vs. 79.4% in GR, p < 0.001), but the specificity did not differ significantly (66.6% vs. 66.4% in BSR, p = 0.238; 70.8% vs. 70.0% in GR, p = 0.689). The average reading time pooled across readers was significantly decreased by AI assistance for BSRs (82.73 vs. 73.04 seconds, p < 0.001) but increased in GRs (35.44 vs. 42.52 seconds, p < 0.001). Conclusion: AI-based software improved the performance of radiologists regardless of their experience and affected the reading time.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.11
• /
• pp.4603-4608
• /
• 2015

Background: To determine the potential value of serum tumor markers in predicting pCR (pathological complete response) during neoadjuvant chemotherapy. Materials and Methods: We retrospectively monitored the pro-, mid-, and post-neoadjuvant treatment serum tumor marker concentrations in patients with locally advanced breast cancer (stage II-III) who accepted pre-surgical chemotherapy or chemotherapy in combination with targeted therapy at Fudan University Shanghai Cancer Center between September 2011 and January 2014 and investigated the association of serum tumor marker levels with therapeutic effect. Core needle biopsy samples were assessed using immunohistochemistry (IHC) prior to neoadjuvant treatment to determine hormone receptor, human epidermal growth factor receptor 2(HER2), and proliferation index Ki67 values. In our study, therapeutic response was evaluated by pCR, defined as the disappearance of all invasive cancer cells from excised tissue (including primary lesion and axillary lymph nodes) after completion of chemotherapy. Analysis of variance of repeated measures and receiver operating characteristic (ROC) curves were employed for statistical analysis of the data. Results: A total of 348 patients were recruited in our study after excluding patients with incomplete clinical information. Of these, 106 patients were observed to have acquired pCR status after treatment completion, accounting for approximately 30.5% of study individuals. In addition, 147patients were determined to be Her-2 positive, among whom the pCR rate was 45.6% (69 patients). General linear model analysis (repeated measures analysis of variance) showed that the concentration of cancer antigen (CA) 15-3 increased after neoadjuvant chemotherapy in both pCR and non-pCR groups, and that there were significant differences between the two groups (P=0.008). The areas under the ROC curves (AUCs) of pre-, mid-, and post-treatment CA15-3 concentrations demonstrated low-level predictive value (AUC=0.594, 0.644, 0.621, respectively). No significant differences in carcinoembryonic antigen (CEA) or CA12-5 serum levels were observed between the pCR and non-pCR groups (P=0.196 and 0.693, respectively). No efficient AUC of CEA or CA12-5 concentrations were observed to predict patient response toward neoadjuvant treatment (both less than 0.7), nor were differences between the two groups observed at different time points. We then analyzed the Her-2 positive subset of our cohort. Significant differences in CEA concentrations were identified between the pCR and non-pCR groups (P=0.039), but not in CA15-3 or CA12-5 levels (p=0.092 and 0.89, respectively). None of the ROC curves showed underlying prognostic value, as the AUCs of these three markers were less than 0.7. The ROC-AUCs for the CA12-5 concentrations of inter-and post-neoadjuvant chemotherapy in the estrogen receptor negative HER2 positive subgroup were 0.735 and 0.767, respectively. However, the specificity and sensitivity values were at odds with each other which meant that improving either the sensitivity or specificity would impair the efficiency of the other. Conclusions: Serum tumor markers CA15-3, CA12-5, and CEA might have little clinical significance in predicting neoadjuvant treatment response in locally advanced breast cancer.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.20 no.9
• /
• pp.868-882
• /
• 2009

In this paper, channel model and wireless link performance analysis for the short-range wireless communication system applications in the terahertz frequency which is currently interested in many countries will be described. In order to realize high data rates above 10 Gbps, the more wide bandwidths will be required than the currently available bandwidths of millimeter-wave frequencies, therefore, the carrier frequencies will be pushed to THz range to obtain larger bandwidths. From the THz atmospheric propagation characteristics based on ITU-R P.676-7, the available bandwidths were calculated to be 68, 48 and 45 GHz at the center frequencies of 220, 300 and 350 GHz, respectively. With these larger bandwidths, it was shown from the simulation that higher data rate above 10 Gbps can be achieved using lower order modulation schemes which have spectral efficiency of below 1. The indoor propagation delay spread characteristics were analyzed using a simplified PDP model with respect to building materials. The RMS delay spread was calculated to be 9.23 ns in a room size of $6\;m(L){\times}5\;m(W){\times}2.5\;m(H)$ for the concrete plaster with TE polarization, which is a similar result of below 10 ns from the Ray-Tracing simulation in the reference paper. The indoor wireless link performance analysis results showed that receiver sensitivity was $-56{\sim}-46\;dBm$ over bandwidth of $5{\sim}50\;GHz$ and antenna gain was calculated to be $26.6{\sim}31.6\;dBi$ at link distance of 10m under the BPSK modulation scheme. The maximum achievable data rates were estimated to be 30, 16 and 12 Gbps at the carrier frequencies of 220, 300 and 350 GHz, respectively, under the A WGN and LOS conditions, where it was assumed that the output power of the transmitter is -15 dBm and link distance of 1 m with BER of $10^{-12}$. If the output power of transmitter is increased, the more higher data rate can be achieved than the above results.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.3
• /
• pp.3-9
• /
• 2017

Recently, in 3rd Generation Partnership Project(3GPP), there is a study of the Long Term Evolution(LTE) based vehicle communication which has been actively conducted to provide a transport efficiency, telematics and infortainment. Because the vehicle communication is closely related to the safety, it requires a reliable communication. Because vehicle speed is very fast, unlike the movement of the user, radio channel is rapidly changed and generate a number of problems such as transmission quality degradation. Therefore, we have to continuously updates the channel estimates. There are five types of conventional channel estimation scheme. Least Square(LS) is obtained by pilot symbol which is known to transmitter and receiver. Decision Directed Channel Estimation(DDCE) scheme uses the data signal for channel estimation. Constructed Data Pilot(CDP) scheme uses the correlation characteristic between adjacent two data symbols. Spectral Temporal Averaging(STA) scheme uses the frequency-time domain average of the channel. Smoothing scheme reduces the peak error value of data decision. In this paper, we propose the novel channel estimation scheme in LTE based Vehicle-to-Vehicle(V2V) environment. In our Hybrid Reliable Channel Estimation(HRCE) scheme, DDCE and Smoothing schemes are combined and finally the Linear Minimum Mean Square Error(LMMSE) scheme is applied to minimize the channel estimation error. Therefore it is possible to detect the reliable data. In simulation results, overall performance can be improved in terms of Normalized Mean Square Error(NMSE) and Bit Error Rate(BER).

• Korean Journal of Health Psychology
• /
• v.14 no.3
• /
• pp.579-596
• /
• 2009

This study aimed to develop the Korean tool of competency to consent to psychiatric treatment and to analyze the reliability and validity of this tool. Also the developed tool's efficiency in determining whether a patient possesses treatment consent competence was checked using the Receiver Operating Characteristic curve and the relevant indices. A total of 193 patients with mental illness, who were hospitalized in a mental hospital or were in community mental health center, participated in this study. We administered a questionnaire consisting of 14 questions concerning understanding, appreciation, reasoning ability, and expression of a choice to the subjects. To investigate the validity of the tool, we conducted the K-MMSE, insight test, estimated IQ, and BPRS. The tool's reliability and usefulness were examined via Cronbach's alpha, ICC, and ROC analysis, and criterion related validation was performed. This tool showed that internal consistency and agreement between raters was relatively high(ICC .80~.98, Cronbach's alpha .56~.83)and the confirmatory factor analysis for constructive validation showed that the tool was valid. Also, estimated IQ, and MMSE were significantly correlated to understanding, appreciation, expression of a choice, and reasoning ability. However, the BPRS did not show significant correlation with any subcompetences. In ROC analysis, full scale cutoff score 18.5 was suggested. Subscale cutoff scores were understanding 4.5, appreciation 8.5, reasoning ability 3.5, and expression of a choice 0.5. These results suggest that this assessment tool is reliable, valid and efficient diagnostically. Finally, limitations and implications of this study were discussed.