Virtual Library

Start Your Search

S. Kumar



Author of

  • +

    P1.08 - Poster Session 1 - Radiotherapy (ID 195)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Radiation Oncology + Radiotherapy
    • Presentations: 1
    • +

      P1.08-006 - Inter-observer variability of GTV delineation based on Lung MRI: impact of radiologist led workshop (ID 779)

      09:30 - 09:30  |  Author(s): S. Kumar

      • Abstract

      Background
      Magnetic Resonance Imaging (MRI) with its superior soft tissue contrast resolution has the potential to improve Gross Tumour Volume (GTV) delineation of lung cancer. The aim of this study was to assess the inter-observer variability between observers for MRI based GTV delineation for Lung cancer and evaluate whether this factor changed following a GTV delineation workshop with a thoracic radiologist.

      Methods
      Five radiation oncologists from three different institutions were asked to delineate GTV on 3 patient datasets. Each observer was given a planning CT, PET, T1 and T2 weighted 1.5 T MRI datasets along with patient history and relevant diagnostic test results. Each observer was instructed to delineate a primary GTV and nodal GTV as required on the T1 and T2 weighted MRI datasets (pre workshop contours). A workshop was then conducted. The aim of the workshop was to discuss each case with a thoracic radiologist and for the radiologist to educate each of the observers in how to review thoracic MRI for both T1 and T2 weighted images. Following the workshop each observer delineated a post workshop GTV. Conformity index (CI) was used to evaluate improvement in inter-observer variability between pre and post workshop contours.

      Results
      Results of two observers are presented here. For patients 1 and 3 slight improvement in CI was noted between pre and post primary and nodal GTV for T1 and T2 weighted datasets. Similarly for patient 3 slight improvements were noted in inter-observer variability for primary GTV for both T1 and T2 weighted images. However there was significant improvement in both T1 and T2 weighted nodal GTV. CI improved from 0.2 to 0.6 and 0-0.6 for T1 and T2 weighted images respectively.

      Conclusion
      Preliminary results from this study indicate that a radiologist led workshop assisted in improving inter-observer variability for MRI based GTV delineation. Further analysis of all observers is required to assess the significance of the impact of a radiologist led contouring workshop in improving inter-observer variability on MRI delineation of lung cancer.

  • +

    P2.08 - Poster Session 2 - Radiotherapy (ID 198)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Radiation Oncology + Radiotherapy
    • Presentations: 2
    • +

      P2.08-011 - Utility of MRI in Lung Cancer radiotherapy: a Literature Review (ID 1342)

      09:30 - 09:30  |  Author(s): S. Kumar

      • Abstract

      Background
      Lung cancer remains a challenging site to treat in radiotherapy. For tumour delineation CT combined with FDG PET is the current gold standard for delineating lung cancer volumes. MRI has been utilized in a limited number of cases such as superior sulcus tumours. However its use has been limited due to reduced MRI signal as a result of low proton density in lung, inhomogeneity of the magnetic field in lung and cardiac and respiratory motion. As technology improves, the utility of MRI in imaging lung cancer has become plausible. The aim of this literature review was to identify evidence to support the use of MRI in lung cancer imaging for radiotherapy.

      Methods
      Pubmed and Google Scholar were used to identify literature on MRI in Lung cancer using the keywords Lung Cancer, MRI, MR, thoracic imaging and radiotherapy. Articles were limited to human and phantom subjects. A total of 22 articles were identified between 1995 and 2013 and reviewed to assess the potential of MRI for lung radiotherapy imaging.

      Results
      Eighteen studies were performed on 1.5T and two on 3T magnets with magnet strength not specified in two studies. Gradient echo sequence and TrueFISP were most common sequences utilised. Thirteen articles examined tumour motion and nine concentrated on tumour volume analysis. One study illustrated minimal geometric distortion and inter-cycle reproducibility of tumour volume on free breathing MRI sequence. The possibility of defining tumour internal target volume for radiotherapy treatment was demonstrated in three studies Two further studies demonstrating variability of lung tumour motion based on location and variability of motion between tumour and non-tumour bearing lung. There was conflicting evidence on pulmonary node detection based on results from two studies. Four comparative studies with FDG PET and functional MRI sequences demonstrated diffusion weighted image (DWI) had higher specificity for lesion detection in the presence of inflammation. Correlation was seen between SUV and DWI value. Two studies demonstrated the potential for radiotherapy planning based on biological function by avoiding functional lung tissue.

      Conclusion
      This literature review indicates that improvement in MRI technology has overcome some of the initial limitations of utilising MRI for lung cancer imaging. MRI can not only provide the morphological information required for identifying lung cancer based on anatomical sequences but also functional information to identify both tumour activity and surrounding healthy or pathological structures. There is potential for MRI to be utilised in the clinical setting for defining tumour volumes for radiotherapy planning and in evaluating tumour response during and after treatment. However further research is required to determine protocols with defined standard sequences specific for lung cancer imaging.

    • +

      P2.08-022 - Non-Small Cell Lung Cancer (NSCLC): Changes in volume during radiotherapy and potential adaptive radiotherapy planning (ID 2511)

      09:30 - 09:30  |  Author(s): S. Kumar

      • Abstract

      Background
      Curative radiotherapy for non-small cell lung cancer (NSCLC) is usually delivered over a time period of 5-7 weeks. Tumour regression has been observed over the course of curative radiotherapy. Currently, the entire radiotherapy treatment is delivered based on the original tumour volume. An evolving approach is adaptive radiotherapy planning whereby the radiotherapy plan is modified during a course of treatment to account for tumour and patient changes. This has the potential to reduce the size of radiotherapy fields, allowing dose escalation and normal tissue sparing.

      Methods
      A cohort of 20 consecutive NSCLC patients receiving a curative course of radiotherapy and who had weekly kV cone beam computer tomography (CBCT) images was identified. The gross tumour volume (GTV) and anatomical reference points were delineated on all CBCT scans. Volume and positional changes were recorded and analyzed.

      Results
      Six patients with non-small cell lung cancer who received curative radiotherapy were assessed so far. Five patients had sufficient image quality for contouring. There was a mean percent decrease of 24.8% by fraction 16 and 37.2% by fraction 26. Average tumour migration was 0.4cm. Progressive anatomical changes were more prominent where there was tumour associated with atelectasis. Updated results for the whole cohort will be presented.

      Conclusion
      Tumour regression was observed in all patients and a proportion showed significant reduction. Adaptive planning was a feasible option in selected patients.

  • +

    P3.08 - Poster Session 3 - Radiotherapy (ID 199)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Radiation Oncology + Radiotherapy
    • Presentations: 1
    • +

      P3.08-010 - The potential use of MRI to delineate lung cancer volumes for radiotherapy (ID 1344)

      09:30 - 09:30  |  Author(s): S. Kumar

      • Abstract

      Background
      The use of MRI for lung cancer volume delineation for radiotherapy is rare. This has been due to poor image quality as a result of physical and physiological factors such as low proton density, susceptibility effects and respiratory and cardiac motion. However as MRI technology has improved, imaging of lung abnormalities has become more feasible. A prospective study was therefore conducted to evaluate image quality for lung tumour delineation on a 1.5T (Tesla) and 3T MRI scanner. The aim of the study was to identify potential scan sequences that could be used clinically for tumour delineation for radiation therapy treatment.

      Methods
      Ten patients with lung cancer underwent MRI, five on a 1.5T GE scanner using a body phased array coil and five on a 3T Phillips scanner. Scans on the 1.5T scanner were undertaken with breath hold and scans on the 3T scanner were performed with respiratory and peripheral pulse gating to give optimal image quality. The thorax was imaged with T2 and T1 weighted sequences on both field strengths. Cine mode imaging to compare tumour motion was also acquired. Scan sequence was matched for the 1.5T and 3T scanners. The quality of images for lung cancer delineation was assessed by an experienced thoracic radiologist and thoracic radiation oncologist using a four point scale. A consensus score ranging from 1(superior) to 4 (inferior) was given for each sequence based on four categories; tumour edge detection, image artefacts, noise affecting edge detection and overall image quality. A score of 2 or below was considered clinically acceptable

      Results
      Both magnet strengths provided reasonable image quality to define tumour volume on lung MRI. The average score for overall image quality between the two scanners was 1.8 for 1.5T and 1.3 for the 3T scanner. For the 1.5T scanner the sagittal and coronal T2 weighted scan scored the best for overall image quality for tumour delineation (1.53), due to limited respiratory motion distortion. However these image planes are not compatible with radiotherapy planning systems. For the 3T scanner the axial T2 images scored best for overall image quality (1.05). For tumour edge detection the sagittal and coronal and T1 weighted images scored best (1.75) for the 1.5T scanner. The axial T2 weighted image and the sagittal cine mode performed best for tumour edge detection (1). Overall sequences on the 3T scanner were rated higher than those on the 1.5T scanner

      Conclusion
      It is feasible to utilise commercially available MRI sequences to acquire images of acceptable quality for the purposes of lung cancer delineation in radiotherapy. Both magnet strengths gave acceptable image quality for clinical use in radiotherapy, with the 3T magnet displaying slightly better image quality. A future study will compare lung cancer delineation between the current standard practice of CT&PET with MRI.