40 ?m in diameter) were accounted for about 15％, and the rest were medium-and small-sized cells.

In the present study the commissural projection between the two superior colliculi in the rat was examined with horseradish peroxidase method. The result shows that when HRP was injected into the superior colliculus of one side, HRP labeled cells could be found in every part of the contralateral superior colliculus. In each case the labeled cells were relatively concentrated in the region corresponding to the site of injection. It indicates that various parts of one superior colliculus may be connected chiefly with the corresponding part of the opposite side through the commissural projection.Most of the labeled cells were found in the middle layer of the superior colliculus, especially in its upper half, less in the deep layer, and the least in the superficial layer. Neurons in the superficial layer of one side project only to the superficial layer of the contralateral side, and so are the middle-deep layers. The connections between the superficial layers of both sides were independent from those of the middledeep layers.The commissural projection of the bilateral superior colliculi passes through the commissure of the superior colliculus which could be divided into a dorsal and a ventral fiberal fasciculns. The dorsal one was smaller, predominantly related to the superficial layer and the upper half of the middle layer; the ventral one was larger, part of its fibers related to the lower half of the middle layer and the deep layer, while the rest project to other nuclei of the contralateral region of the mesencephalon (e. g. nucleus cuneiforms, etc,).Most of the labeled cells were small in size, the rest were medium-sized, and no large ones were found.

It is demonstrated that the parabigeminal nucleus of the rat is subdivided intodorsal,middle and ventral groups.The parabigeminal nucleus sends fibers to bila-teral superior colliculus,the rostral half of it predominantly projects to the rostralhalf of the contralateral superior colliculus,but the rostral end projects only to thecontralateral one;the caudal half of this nucleus predominantly projects to thecaudal half of the ipsilateral superior colliculus,but the caudal end projects only tothe ipsilateral one.The superficial layer of the superior colliculus receives projec-tions from the ipsilateral dorsal and ventral groups and from the contralateralmiddle group of the parabigeminal nucleus.The middle and deep layers receive pro-jections from the ipsilateral middle group and contralateral dorsal and ventral groups,and probably from the other groups of both sides.The superior colliculus also sends fibers to both parabigeminal nuclei,predo-minantly the ipsilateral side.The lateral tegmental area sends fibers to the middle and deep layers of the su-perior colliculus.From the results described above,it could be concluded that the parabigeminalnucleus——tectum——parabigeminal nucleus connections are not only concerned withsuperficial layer,but also with the middle and deep layers of the superior colliculus.The tegmentum——tectum——tegmentum connections are only concerned with the mid-dle and deep layers of the superior colliculus.

Neurons descending from the griseum centrale mesencephali, nucleus Darkschewitsch and nucleus interstitialis of Cajal to the nucleus raphe magnus and adjacent reticular formation (nucleus reticularis gigantocellularis and nucleus reticularis pontis caudalis) were identified in 9 adult cats with the retrograde HRP method. In the griseum centrale mesencephali, the labeled neurons were found bilaterally but slightly more ipsilaterally. In the nucleus Darkschewitscb and nucleus interstitialis of Cajal, the labeled neurons were consistently found in its rostral part ipsilateral to the injected side at the level of the posterior commissure. In addition, in 5 of the 9 cases, a few labeled neurons were observed in the nucleus raphe dorsalis.

The projections of diencephalon onto the rostral part of the mesencephalic periaqueductal gray matter (PAG) were investigated by using the method of retrograde transport of HRP in the rat. The results have led to following conclusions:1. The diencephalic afferents to the rostral part of PAG originate mainly from medial preoptic area, anterior hypothalamic area, dorsomedial and ventromedial hypothalamic nuclei, dorsal premamillary nucleus, lateral hypothalamic region, Forel's field and zona incerta ipsilaterally, and a few of them originate contralaterally.2. The anterior hypothalamic area, ventromedial hypothalamic nucleus, dorsal premamillary nucleus and zona incerta project more to the ipsilateral nucleus: lateralis of PAG than to the nucleus medialis and contralateral nucleus lateralis of PAG.3. In the anterior hypothalamic area, a large number of neurons projecting to PAG was located in the lateral anterior nucleus.4. In the lateral hypothalamic region, the lateral preoptic area and anterior division project to both nuclei lateralis and medialis, while its tuberal and mamillary divisions probably project only to nucleus lateralis.

In the present study ~3H-Leucine or WGA-HRP was injected into the superior colliculus of one side in the rat.The terminal areas of the efferent projection fibers from the superior colliculus were examined. The efferent fibers of the superfical layer of the superior colliculus descended ipsilaterally to terminate in the parabigeminal nucleus(predominantly the dorsal and ventral part of the same side)and dorso-lateral part of the ipsilateral pontine nucleus.Ascending projections terminated to the medial geniculate nucleus,the posterior pretectal nucleus and latero-posterior nucleus of the thalamus(all bila- terally,but with ipsilateral predominance),the ipsilateral medial and lateral optic nuclei,and the dorsal and ventral lateral geniculate nucleus.In addition,labeled granules were also found in bilateral optic tracts and the optic chiasma. The efferent fibers of the middle and deep layers terminated to the ipsilateral central gray,the nucleus of Darkschewitsch,the interstitial nucleus of Cajal,the cuneiform nucleus and the contralateral superior colliculus.Ascending fibers ter- minated to the medial geniculate nucleus,the suprageniculate nucleus,the anterior- pretectal nucleus,the postero-lateral nucleus of the thalamus(all bilaterally,but more on the ipsilateral side),the parafascicular nucleus,the zona incerta,the ventral nucleus of the thalamus(all ipsilaterally).Descending fibers terminated to ipsilateral parabigeminal area and the parabigeminal nucleus,the dorso-lateral part of the pontine nucleus,the lateral part of the inferior colliculus,the reticular formation of the medulla oblongata and pons,and the lateral part of the inferior olive.The fibers terminated also to contralateral nuclei such as the parabigeminal nucleus,the medial part of the reticular formation of the medulla oblongata and ports,the medial accessory nucleus of the inferior olive,the anterior horn of the cervical spinal cord.

Objective:To comprehensively evaluate the effects of different bolus usages in postmastectomy intensity-modulated radiotherapy (PM-IMRT) on doses.Methods:Fifty patients receiving PM-IMRT at Fudan University Shanghai Cancer Center from April to October 2021 were retrospectively studied. The planning target volume (PTV) was divided into four parts, namely chest wall (CW), internal mammary node, retained axillary lymph node, and supraclavicular node. The prescription dose was 50 Gy/25 fractions. Three PM-IMRT plans applying boluses with different thicknesses (3, 5 and 10 mm) were designed for each patient. The effects of different thicknesses and usage frequencies of boluses on PTV coverage, high dose volume of the CW skin, and dose to surrounding normal tissues were comprehensively evaluated.Results:When boluses were applied throughout the PM-IMRT, the PTV V95% of plans applying 10 mm-thick boluses was lower than that of plans applying 3 and 5 mm-thick boluses ( F=3.340, P < 0.05), the CI of plans applying 3 mm-thick boluses was higher than that of plans applying 5 and 10 mm-thick boluses ( F = 50.05, P < 0.05), and there was no statistically significant differences in the skin V105% and V110% of three plans( P > 0.05). Both PTV V95% and skin V105% were reduced with a decrease in the usage frequency of boluses. At a frequency of 20, PTV V95% decreased slightly (< 1%), while skin V105% decreased sharply to nearly half of the original values. At a frequency of 15, the PTV V95%, CI, and HI in the three plans showed no statistically significant dosimetric differences ( P > 0.05). The PTV Dmax of plans applying 3 mm-thick boluses was lower than that of plans applying 5 and 10 mm-thick boluses ( F = 9.21, P < 0.05). As for the dose to surrounding normal tissues, different bolus thicknesses and frequencies had negligible effects on doses to heart and lung, causing little different biological effects. Conclusions:For PM-IMRT, different bolus thicknesses have similar effects on doses to the PTV, skin, heart, and lung. Bolus usage frequency, rather than thickness, was the major factor determining the PTV coverage and the dose to CW skin.

Objective To explore the importance of head fixation in chest wall field combined with supraclavicular field radiotherapy for breast cancer by comparing the displacement error and dosimetric differences caused by multi-functional body board and breast bracket.Methods Thirty patients with breast cancer were randomly divided into groups A and B.In group A,patients were fixed with multi-functional body board and head thermoplastic film.In group B,patients were fixed with traditional breast brackets.Each patient received CBCT scan before and after radiotherapy.Both setup errors and intra-fractional displacements in the x-,y-and z-axis,V100 and V95 were calculated.Statistical analyses were performed using the independent sample t-test.Results The displacement errors in groups A and B before and after radiotherapy were (1.24± 0.42),(1.71± 0.61) and (2.25± 1.04) mm vs.(3.67± 2.05),(3.78± 1.74),(4.65±2.66) mm in the x-,y-and z-axis,respectively (P=0.033,0.027,0.020).The intra-fractional displacements in groups A and B were (1.10±0.66),(1.13±0.59),(1.11 ±0.62) mm vs.(2.48±0.88),(2.21 ±0.98),(3.53±2.01) mm in the x-,y-and z-axis,respectively (P=0.030,0.021,0.013).The V100 in groups A and B were (94.27± 3.20) ％ and (99.08± 0.60) ％ (P =0.065),and (89.48± 4.70) ％ and (96.53± 2.50) ％ for V95 (P =0.002),respectively.Conclusion The risk of displacement error is significantly reduced using multi-functional body board,which enhances the accuracy of radiation dose in chest wall and supraclavicular fields of breast cancer patients.