Evaluation of an artificial intelligence U-net algorithm for pulmonary nodule tracking on chest computed tomography images.

OBJECTIVES: To apply image registration in the follow up of lung nodules and verify the feasibility of automatic tracking of lung nodules using an artificial intelligence (AI) method. METHODS: For this retrospective, observational study, patients with pulmonary nodules 5-30 mm in diameter on computed tomography (CT) and who had at least six months follow-up were identified. Two radiologists defined a 'correct' cuboid circumscribing each nodule which was used to judge the success/failure of nodule tracking. An AI algorithm was applied in which a U-net type neural network model was trained to predict the deformation vector field between two examinations. When the estimated position was within a defined cuboid, the AI algorithm was judged a success. RESULTS: In total, 49 lung nodules in 40 patients, with a total of 368 follow-up CT examinations were examined. The success rate for each time evaluation was 94% (345/368) and for 'nodule-by-nodule evaluation' was 78% (38/49). Reasons for a decrease in success rate were related to small nodules and those that decreased in size. CONCLUSION: Automatic tracking of lung nodules is highly feasible.

Feasibility of noise-reduction reconstruction technology based on non-local-mean principle in SiPM-PET/CT.

Quantitative values of positron emission tomography (PET) images using non-local-mean in a silicon photomultiplier (SiPM)-PET/computed tomography (CT) system with phantom and clinical images. The evaluation was conducted on a National Electrical Manufacturers Association body phantom with micro-spheres (4, 5, 6, 8, 10, 13 mm) and clinical images using the SiPM-PET/CT system. The signal-to-background ratio of the phantom was set to 4, and all PET image data was obtained and reconstructed using three-dimensional ordered subset expectation maximization, time-of-flight, point-spread function, and a 4-mm Gaussian filter (GF) and clear adaptive low-noise method (CaLM) in mild, standard, and strong intensities. The evaluation included the standardized uptake value (SUV), percent contrast (QH), coefficient of variation of the background area (CVbackground) clinical imaging for SUV of lung nodules, liver signal-to-noise ratio (SNR), and visual evaluation. SUVmax for 8-mm sphere in phantom images at 2 min for GF and CaLM (mild, standard, strong) were 2.11, 2.32, 2.02, and 1.72; the QH, 8 mm was 27.33 %, 27.47 %, 21.81 %, and 16.09 %; and CVbackground was 12.78, 11.35, 7.86, and 4.71, respectively. CaLM demonstrated higher SUVmax in clinical images than GF for all lung nodule sizes. The average SUVmax for nodules with a diameter of ≤ 1 cm were 5.9 ± 2.4, 9.9 ± 4.9, 9.9 ± 5.0, and 9.9 ± 5.0 for GF and CaLM-mild, standard, and strong intensities, respectively. Liver SNRs were higher for CaLM (mild, standard, strong) compared to GF, with increasing CaLM intensity causing higher liver SNR. CaLM-mild and standard demonstrated suitability for diagnosis in visual evaluation.

Impact of list-mode reconstruction and image-space point spread function correction on PET image contrast and quantitative value using SiPM-based PET/CT system.

We evaluate the effects of list-mode reconstruction and the image-space point spread function (iPSF) on the contrast and quantitative values of positron emission tomography (PET) images using a SiPM-PET/CT system. The evaluation is conducted on an NEMA body phantom and clinical images using a Cartesion Prime SiPM-PET/CT system. The signal-to-background ratio (SBR) of the phantom is set to 2, 4, 6, and 8, and all the PET image data are obtained and reconstructed using 3D-OSEM, time-of-flight, iPSF (-/ +), and a 4-mm Gaussian filter with several iterations. The evaluation criteria include % background variability (NB,10 mm), % contrast (QH,10 mm), iPSF change in QH,10 mm (ΔQH,10 mm) for edge artifact evaluation, profile curves, visual evaluation of edge artifacts, clinical imaging for the standardized uptake value (SUV) of lung nodules, and SNRliver. NB,10 mm demonstrates no significant difference in all SBRs with and without iPSF, whereas QH,10 mm is higher based on the SBR with and without iPSF. ΔQH,10 mm indicates increased iterations and a larger rate of change (> 5%) for small spheres of < 17 mm. The profile curves portrayed almost real concentrations, except for the 10-mm sphere of SBR2 without iPSF; however, with iPSF, an overshoot was observed in the 13-mm sphere of all SBRs. The degree of overshoot increased with increasing iteration and SBR. Edge artifacts were detected at values ≥ 17-22 mm in SBRs other than SBR2 with iPSF. Irrespective of the nodal size, SUV and SNRliver improved considerably after iPSF adjustment. Therefore, the effects of list-mode reconstruction and iPSF on PET image contrast were limited, and the overcorrection of the quantitative values was validated using iPSF.

A Time-of-Flight Image Sensor Using 8-Tap P-N Junction Demodulator Pixels.

This paper presents a time-of-flight image sensor based on 8-Tap P-N junction demodulator (PND) pixels, which is designed for hybrid-type short-pulse (SP)-based ToF measurements under strong ambient light. The 8-tap demodulator implemented with multiple p-n junctions used for modulating the electric potential to transfer photoelectrons to eight charge-sensing nodes and charge drains has an advantage of high-speed demodulation in large photosensitive areas. The ToF image sensor implemented using 0.11 µm CIS technology, consisting of an 120 (H) × 60 (V) image array of the 8-tap PND pixels, successfully works with eight consecutive time-gating windows with the gating width of 10 ns and demonstrates for the first time that long-range (>10 m) ToF measurements under high ambient light are realized using single-frame signals only, which is essential for motion-artifact-free ToF measurements. This paper also presents an improved depth-adaptive time-gating-number assignment (DATA) technique for extending the depth range while having ambient-light canceling capability and a nonlinearity error correction technique. By applying these techniques to the implemented image sensor chip, hybrid-type single-frame ToF measurements with depth precision of maximally 16.4 cm (1.4% of the maximum range) and the maximum non-linearity error of 0.6% for the full-scale depth range of 1.0-11.5 m and operations under direct-sunlight-level ambient light (80 klux) have been realized. The depth linearity achieved in this work is 2.5 times better than that of the state-of-the-art 4-tap hybrid-type ToF image sensor.

Development and performance evaluation of a deep learning lung nodule detection system.

BACKGROUND: Lung cancer is the leading cause of cancer-related deaths throughout the world. Chest computed tomography (CT) is now widely used in the screening and diagnosis of lung cancer due to its effectiveness. Radiologists must identify each small nodule shadow from 3D volume images, which is very burdensome and often results in missed nodules. To address these challenges, we developed a computer-aided detection (CAD) system that automatically detects lung nodules in CT images. METHODS: A total of 1997 chest CT scans were collected for algorithm development. The algorithm was designed using deep learning technology. In addition to evaluating detection performance on various public datasets, its robustness to changes in radiation dose was assessed by a phantom study. To investigate the clinical usefulness of the CAD system, a reader study was conducted with 10 doctors, including inexperienced and expert readers. This study investigated whether the use of the CAD as a second reader could prevent nodular lesions in lungs that require follow-up examinations from being overlooked. Analysis was performed using the Jackknife Free-Response Receiver-Operating Characteristic (JAFROC). RESULTS: The CAD system achieved sensitivity of 0.98/0.96 at 3.1/7.25 false positives per case on two public datasets. Sensitivity did not change within the range of practical doses for a study using a phantom. A second reader study showed that the use of this system significantly improved the detection ability of nodules that could be picked up clinically (p = 0.026). CONCLUSIONS: We developed a deep learning-based CAD system that is robust to imaging conditions. Using this system as a second reader increased detection performance.

An 8-Tap CMOS Lock-In Pixel Image Sensor for Short-Pulse Time-of-Flight Measurements.

An 8-tap CMOS lock-in pixel image sensor that has seven carrier-capturing and a draining time window was developed for short-pulse time-of-flight (TOF) measurements. The proposed pixel for the short-pulse TOF measurements has seven consecutive time-gating windows, each of which has the width of 6 ns, which is advantageous for high-resolution range imaging, particularly for relatively longer distances (>5 m) and under high ambient light operations. In order to enhance the depth resolution, a technique for the depth-adaptive time-gating-number assignment (DATA) for the short-pulse TOF measurement is proposed. A prototype of the 8-tap CMOS lock-in pixel image sensor is implemented with a 1POLY 4METAL 0.11-µm CIS process. The maximum non-linearity error of 1.56%FS for the range of 1-6.4 m and the depth resolution of 6.4 mm was obtained at 6.2 m using the DATA technique.

[Optimization of Scan Parameters for Patient without Breath Hold in Chest by Computed Tomography].

This study aims to establish optimal scan parameters by high temporal resolution computed tomography (CT) scan for emergency patients who cannot hold their breath. First, we investigated scan parameters that can reduce the effect of motion by evaluating motion artifacts from the moving phantom scan and the temporal sensitivity profile (TPS) measurement. Second, we confirmed the standard deviation (SD) of the CT values as well as the operating time and exposure dose. As the results, plan C [rotation time: 0.275 s, detector rows: 80, pitch factor (PF): 1.100] and plan E (rotation time: 0.275 s, detector rows: 100, PF: 0.880) demonstrated high temporal resolution. The difference between the two is PF. The noise of plan C increased because PF is higher than plan E. This is also evident from the results of SD measurement. Our study demonstrates that the optimal parameters for patients who cannot hold their breath in the emergency care are plan C and plan E. In conclusion, we clarified necessary optimal scan parameters to provide clinical image that has more diagnostic information by reducing the effect of breath motion for emergency patients.